diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17813 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.5048065072713825,
+  "eval_steps": 500,
+  "global_step": 2540,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009859502095144195,
+      "grad_norm": 1.3765249905290702,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8092,
+      "step": 1
+    },
+    {
+      "epoch": 0.001971900419028839,
+      "grad_norm": 1.367289057075295,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8,
+      "step": 2
+    },
+    {
+      "epoch": 0.0029578506285432584,
+      "grad_norm": 1.3431155113747304,
+      "learning_rate": 6e-06,
+      "loss": 0.8049,
+      "step": 3
+    },
+    {
+      "epoch": 0.003943800838057678,
+      "grad_norm": 1.151248103486504,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8039,
+      "step": 4
+    },
+    {
+      "epoch": 0.004929751047572098,
+      "grad_norm": 0.8104674792888513,
+      "learning_rate": 1e-05,
+      "loss": 0.7477,
+      "step": 5
+    },
+    {
+      "epoch": 0.005915701257086517,
+      "grad_norm": 2.727064547227183,
+      "learning_rate": 9.999997324837724e-06,
+      "loss": 0.7947,
+      "step": 6
+    },
+    {
+      "epoch": 0.006901651466600937,
+      "grad_norm": 1.0749417778668735,
+      "learning_rate": 9.99998929935376e-06,
+      "loss": 0.7307,
+      "step": 7
+    },
+    {
+      "epoch": 0.007887601676115356,
+      "grad_norm": 0.7628530475064141,
+      "learning_rate": 9.999975923556696e-06,
+      "loss": 0.7165,
+      "step": 8
+    },
+    {
+      "epoch": 0.008873551885629776,
+      "grad_norm": 0.5857783129273212,
+      "learning_rate": 9.999957197460844e-06,
+      "loss": 0.73,
+      "step": 9
+    },
+    {
+      "epoch": 0.009859502095144195,
+      "grad_norm": 0.5407067952150867,
+      "learning_rate": 9.99993312108624e-06,
+      "loss": 0.6877,
+      "step": 10
+    },
+    {
+      "epoch": 0.010845452304658615,
+      "grad_norm": 0.39622766885472044,
+      "learning_rate": 9.999903694458653e-06,
+      "loss": 0.6758,
+      "step": 11
+    },
+    {
+      "epoch": 0.011831402514173034,
+      "grad_norm": 0.3931749811712902,
+      "learning_rate": 9.999868917609565e-06,
+      "loss": 0.6651,
+      "step": 12
+    },
+    {
+      "epoch": 0.012817352723687454,
+      "grad_norm": 0.40125124272709395,
+      "learning_rate": 9.999828790576194e-06,
+      "loss": 0.6986,
+      "step": 13
+    },
+    {
+      "epoch": 0.013803302933201873,
+      "grad_norm": 0.3370162547077992,
+      "learning_rate": 9.999783313401478e-06,
+      "loss": 0.6799,
+      "step": 14
+    },
+    {
+      "epoch": 0.014789253142716292,
+      "grad_norm": 0.2931745441380535,
+      "learning_rate": 9.999732486134078e-06,
+      "loss": 0.6626,
+      "step": 15
+    },
+    {
+      "epoch": 0.01577520335223071,
+      "grad_norm": 0.33086761396823533,
+      "learning_rate": 9.999676308828385e-06,
+      "loss": 0.6753,
+      "step": 16
+    },
+    {
+      "epoch": 0.01676115356174513,
+      "grad_norm": 0.36273495900838076,
+      "learning_rate": 9.999614781544512e-06,
+      "loss": 0.6809,
+      "step": 17
+    },
+    {
+      "epoch": 0.017747103771259553,
+      "grad_norm": 0.3269646639666865,
+      "learning_rate": 9.999547904348294e-06,
+      "loss": 0.6486,
+      "step": 18
+    },
+    {
+      "epoch": 0.018733053980773972,
+      "grad_norm": 0.311315845954239,
+      "learning_rate": 9.999475677311298e-06,
+      "loss": 0.6707,
+      "step": 19
+    },
+    {
+      "epoch": 0.01971900419028839,
+      "grad_norm": 0.2646645447167399,
+      "learning_rate": 9.99939810051081e-06,
+      "loss": 0.6517,
+      "step": 20
+    },
+    {
+      "epoch": 0.02070495439980281,
+      "grad_norm": 0.24097098469055522,
+      "learning_rate": 9.999315174029843e-06,
+      "loss": 0.6188,
+      "step": 21
+    },
+    {
+      "epoch": 0.02169090460931723,
+      "grad_norm": 0.2715478883936797,
+      "learning_rate": 9.999226897957132e-06,
+      "loss": 0.6572,
+      "step": 22
+    },
+    {
+      "epoch": 0.022676854818831648,
+      "grad_norm": 0.2542812528029893,
+      "learning_rate": 9.99913327238714e-06,
+      "loss": 0.652,
+      "step": 23
+    },
+    {
+      "epoch": 0.023662805028346067,
+      "grad_norm": 0.2569805794182414,
+      "learning_rate": 9.999034297420053e-06,
+      "loss": 0.6522,
+      "step": 24
+    },
+    {
+      "epoch": 0.02464875523786049,
+      "grad_norm": 0.2396967420842906,
+      "learning_rate": 9.998929973161777e-06,
+      "loss": 0.635,
+      "step": 25
+    },
+    {
+      "epoch": 0.02563470544737491,
+      "grad_norm": 0.24159797374059683,
+      "learning_rate": 9.99882029972395e-06,
+      "loss": 0.6372,
+      "step": 26
+    },
+    {
+      "epoch": 0.026620655656889328,
+      "grad_norm": 0.2519427485321024,
+      "learning_rate": 9.998705277223926e-06,
+      "loss": 0.6374,
+      "step": 27
+    },
+    {
+      "epoch": 0.027606605866403747,
+      "grad_norm": 0.23714310942142935,
+      "learning_rate": 9.99858490578479e-06,
+      "loss": 0.6304,
+      "step": 28
+    },
+    {
+      "epoch": 0.028592556075918166,
+      "grad_norm": 0.23852507363196584,
+      "learning_rate": 9.998459185535342e-06,
+      "loss": 0.6167,
+      "step": 29
+    },
+    {
+      "epoch": 0.029578506285432585,
+      "grad_norm": 0.2237646708710333,
+      "learning_rate": 9.998328116610118e-06,
+      "loss": 0.6265,
+      "step": 30
+    },
+    {
+      "epoch": 0.030564456494947004,
+      "grad_norm": 0.22617135505726307,
+      "learning_rate": 9.998191699149367e-06,
+      "loss": 0.6232,
+      "step": 31
+    },
+    {
+      "epoch": 0.03155040670446142,
+      "grad_norm": 0.2367290677204515,
+      "learning_rate": 9.99804993329906e-06,
+      "loss": 0.6334,
+      "step": 32
+    },
+    {
+      "epoch": 0.03253635691397584,
+      "grad_norm": 0.23223879565221392,
+      "learning_rate": 9.997902819210903e-06,
+      "loss": 0.6479,
+      "step": 33
+    },
+    {
+      "epoch": 0.03352230712349026,
+      "grad_norm": 0.21858220436116263,
+      "learning_rate": 9.997750357042315e-06,
+      "loss": 0.6214,
+      "step": 34
+    },
+    {
+      "epoch": 0.03450825733300468,
+      "grad_norm": 0.20343274643065565,
+      "learning_rate": 9.997592546956439e-06,
+      "loss": 0.6142,
+      "step": 35
+    },
+    {
+      "epoch": 0.035494207542519106,
+      "grad_norm": 0.2577948288273245,
+      "learning_rate": 9.997429389122141e-06,
+      "loss": 0.6379,
+      "step": 36
+    },
+    {
+      "epoch": 0.036480157752033525,
+      "grad_norm": 0.20636422243489688,
+      "learning_rate": 9.997260883714015e-06,
+      "loss": 0.6348,
+      "step": 37
+    },
+    {
+      "epoch": 0.037466107961547944,
+      "grad_norm": 0.20502145548371767,
+      "learning_rate": 9.99708703091237e-06,
+      "loss": 0.6065,
+      "step": 38
+    },
+    {
+      "epoch": 0.03845205817106236,
+      "grad_norm": 0.21299232506890886,
+      "learning_rate": 9.996907830903238e-06,
+      "loss": 0.6263,
+      "step": 39
+    },
+    {
+      "epoch": 0.03943800838057678,
+      "grad_norm": 0.20706080631505255,
+      "learning_rate": 9.996723283878376e-06,
+      "loss": 0.6138,
+      "step": 40
+    },
+    {
+      "epoch": 0.0404239585900912,
+      "grad_norm": 0.21574359503465926,
+      "learning_rate": 9.996533390035264e-06,
+      "loss": 0.6359,
+      "step": 41
+    },
+    {
+      "epoch": 0.04140990879960562,
+      "grad_norm": 0.2094884886807774,
+      "learning_rate": 9.996338149577098e-06,
+      "loss": 0.6233,
+      "step": 42
+    },
+    {
+      "epoch": 0.04239585900912004,
+      "grad_norm": 0.21797785845613704,
+      "learning_rate": 9.996137562712798e-06,
+      "loss": 0.6074,
+      "step": 43
+    },
+    {
+      "epoch": 0.04338180921863446,
+      "grad_norm": 0.20885248816596044,
+      "learning_rate": 9.995931629657005e-06,
+      "loss": 0.609,
+      "step": 44
+    },
+    {
+      "epoch": 0.04436775942814888,
+      "grad_norm": 0.2300394406472479,
+      "learning_rate": 9.995720350630083e-06,
+      "loss": 0.6222,
+      "step": 45
+    },
+    {
+      "epoch": 0.045353709637663296,
+      "grad_norm": 0.2083396663165689,
+      "learning_rate": 9.99550372585811e-06,
+      "loss": 0.6122,
+      "step": 46
+    },
+    {
+      "epoch": 0.046339659847177715,
+      "grad_norm": 0.20433334433389508,
+      "learning_rate": 9.995281755572891e-06,
+      "loss": 0.6177,
+      "step": 47
+    },
+    {
+      "epoch": 0.047325610056692134,
+      "grad_norm": 0.1973460676616573,
+      "learning_rate": 9.99505444001195e-06,
+      "loss": 0.6073,
+      "step": 48
+    },
+    {
+      "epoch": 0.04831156026620655,
+      "grad_norm": 0.2562590619272977,
+      "learning_rate": 9.994821779418529e-06,
+      "loss": 0.5842,
+      "step": 49
+    },
+    {
+      "epoch": 0.04929751047572098,
+      "grad_norm": 0.1988436425039502,
+      "learning_rate": 9.994583774041588e-06,
+      "loss": 0.6125,
+      "step": 50
+    },
+    {
+      "epoch": 0.0502834606852354,
+      "grad_norm": 0.21977567705232728,
+      "learning_rate": 9.994340424135808e-06,
+      "loss": 0.6044,
+      "step": 51
+    },
+    {
+      "epoch": 0.05126941089474982,
+      "grad_norm": 0.1940896134389713,
+      "learning_rate": 9.994091729961593e-06,
+      "loss": 0.6047,
+      "step": 52
+    },
+    {
+      "epoch": 0.052255361104264236,
+      "grad_norm": 0.19863138435651861,
+      "learning_rate": 9.993837691785058e-06,
+      "loss": 0.6192,
+      "step": 53
+    },
+    {
+      "epoch": 0.053241311313778655,
+      "grad_norm": 0.26855885840780924,
+      "learning_rate": 9.99357830987804e-06,
+      "loss": 0.613,
+      "step": 54
+    },
+    {
+      "epoch": 0.054227261523293074,
+      "grad_norm": 0.2004024021114952,
+      "learning_rate": 9.9933135845181e-06,
+      "loss": 0.6067,
+      "step": 55
+    },
+    {
+      "epoch": 0.05521321173280749,
+      "grad_norm": 0.21801464983753077,
+      "learning_rate": 9.993043515988504e-06,
+      "loss": 0.6115,
+      "step": 56
+    },
+    {
+      "epoch": 0.05619916194232191,
+      "grad_norm": 0.19078374812598073,
+      "learning_rate": 9.992768104578248e-06,
+      "loss": 0.5975,
+      "step": 57
+    },
+    {
+      "epoch": 0.05718511215183633,
+      "grad_norm": 0.19814776870710335,
+      "learning_rate": 9.992487350582037e-06,
+      "loss": 0.5711,
+      "step": 58
+    },
+    {
+      "epoch": 0.05817106236135075,
+      "grad_norm": 0.20847307806857995,
+      "learning_rate": 9.992201254300299e-06,
+      "loss": 0.6151,
+      "step": 59
+    },
+    {
+      "epoch": 0.05915701257086517,
+      "grad_norm": 0.18634768164717536,
+      "learning_rate": 9.991909816039174e-06,
+      "loss": 0.5866,
+      "step": 60
+    },
+    {
+      "epoch": 0.06014296278037959,
+      "grad_norm": 0.22058492402253121,
+      "learning_rate": 9.991613036110517e-06,
+      "loss": 0.6051,
+      "step": 61
+    },
+    {
+      "epoch": 0.06112891298989401,
+      "grad_norm": 0.19257108247180021,
+      "learning_rate": 9.991310914831908e-06,
+      "loss": 0.6024,
+      "step": 62
+    },
+    {
+      "epoch": 0.06211486319940843,
+      "grad_norm": 0.196142910088349,
+      "learning_rate": 9.991003452526632e-06,
+      "loss": 0.6024,
+      "step": 63
+    },
+    {
+      "epoch": 0.06310081340892285,
+      "grad_norm": 0.20401755542944414,
+      "learning_rate": 9.990690649523694e-06,
+      "loss": 0.6013,
+      "step": 64
+    },
+    {
+      "epoch": 0.06408676361843726,
+      "grad_norm": 0.19636579730374507,
+      "learning_rate": 9.990372506157813e-06,
+      "loss": 0.5949,
+      "step": 65
+    },
+    {
+      "epoch": 0.06507271382795168,
+      "grad_norm": 0.1848833418382152,
+      "learning_rate": 9.990049022769426e-06,
+      "loss": 0.591,
+      "step": 66
+    },
+    {
+      "epoch": 0.0660586640374661,
+      "grad_norm": 0.19274419278695507,
+      "learning_rate": 9.989720199704678e-06,
+      "loss": 0.5972,
+      "step": 67
+    },
+    {
+      "epoch": 0.06704461424698052,
+      "grad_norm": 0.1980938095739198,
+      "learning_rate": 9.989386037315433e-06,
+      "loss": 0.5877,
+      "step": 68
+    },
+    {
+      "epoch": 0.06803056445649494,
+      "grad_norm": 0.19027491755499437,
+      "learning_rate": 9.989046535959269e-06,
+      "loss": 0.5849,
+      "step": 69
+    },
+    {
+      "epoch": 0.06901651466600936,
+      "grad_norm": 0.20011979699788518,
+      "learning_rate": 9.988701695999467e-06,
+      "loss": 0.6015,
+      "step": 70
+    },
+    {
+      "epoch": 0.07000246487552379,
+      "grad_norm": 0.19268639996801692,
+      "learning_rate": 9.988351517805034e-06,
+      "loss": 0.6134,
+      "step": 71
+    },
+    {
+      "epoch": 0.07098841508503821,
+      "grad_norm": 0.19634572510337903,
+      "learning_rate": 9.987996001750682e-06,
+      "loss": 0.5951,
+      "step": 72
+    },
+    {
+      "epoch": 0.07197436529455263,
+      "grad_norm": 0.19629312560061116,
+      "learning_rate": 9.987635148216837e-06,
+      "loss": 0.6058,
+      "step": 73
+    },
+    {
+      "epoch": 0.07296031550406705,
+      "grad_norm": 0.20573948484183535,
+      "learning_rate": 9.987268957589633e-06,
+      "loss": 0.6016,
+      "step": 74
+    },
+    {
+      "epoch": 0.07394626571358147,
+      "grad_norm": 0.20941911620613557,
+      "learning_rate": 9.986897430260922e-06,
+      "loss": 0.5855,
+      "step": 75
+    },
+    {
+      "epoch": 0.07493221592309589,
+      "grad_norm": 0.19813441493160366,
+      "learning_rate": 9.986520566628256e-06,
+      "loss": 0.5926,
+      "step": 76
+    },
+    {
+      "epoch": 0.0759181661326103,
+      "grad_norm": 0.19873514925053815,
+      "learning_rate": 9.986138367094913e-06,
+      "loss": 0.5963,
+      "step": 77
+    },
+    {
+      "epoch": 0.07690411634212473,
+      "grad_norm": 0.18811627969165587,
+      "learning_rate": 9.985750832069861e-06,
+      "loss": 0.5815,
+      "step": 78
+    },
+    {
+      "epoch": 0.07789006655163914,
+      "grad_norm": 0.19234106290642222,
+      "learning_rate": 9.985357961967795e-06,
+      "loss": 0.5949,
+      "step": 79
+    },
+    {
+      "epoch": 0.07887601676115356,
+      "grad_norm": 0.20955249698840359,
+      "learning_rate": 9.984959757209108e-06,
+      "loss": 0.5877,
+      "step": 80
+    },
+    {
+      "epoch": 0.07986196697066798,
+      "grad_norm": 0.19383207007852313,
+      "learning_rate": 9.984556218219908e-06,
+      "loss": 0.5966,
+      "step": 81
+    },
+    {
+      "epoch": 0.0808479171801824,
+      "grad_norm": 0.1947234699228627,
+      "learning_rate": 9.984147345432003e-06,
+      "loss": 0.5768,
+      "step": 82
+    },
+    {
+      "epoch": 0.08183386738969682,
+      "grad_norm": 0.19783757001989402,
+      "learning_rate": 9.983733139282917e-06,
+      "loss": 0.5843,
+      "step": 83
+    },
+    {
+      "epoch": 0.08281981759921124,
+      "grad_norm": 0.21204862645501593,
+      "learning_rate": 9.983313600215876e-06,
+      "loss": 0.6108,
+      "step": 84
+    },
+    {
+      "epoch": 0.08380576780872566,
+      "grad_norm": 0.18979667022849442,
+      "learning_rate": 9.982888728679816e-06,
+      "loss": 0.586,
+      "step": 85
+    },
+    {
+      "epoch": 0.08479171801824008,
+      "grad_norm": 0.35843719108760985,
+      "learning_rate": 9.982458525129377e-06,
+      "loss": 0.5756,
+      "step": 86
+    },
+    {
+      "epoch": 0.0857776682277545,
+      "grad_norm": 0.2017647897675324,
+      "learning_rate": 9.982022990024903e-06,
+      "loss": 0.6062,
+      "step": 87
+    },
+    {
+      "epoch": 0.08676361843726892,
+      "grad_norm": 0.18829410076504116,
+      "learning_rate": 9.981582123832443e-06,
+      "loss": 0.583,
+      "step": 88
+    },
+    {
+      "epoch": 0.08774956864678334,
+      "grad_norm": 0.20926053875571596,
+      "learning_rate": 9.981135927023758e-06,
+      "loss": 0.5939,
+      "step": 89
+    },
+    {
+      "epoch": 0.08873551885629775,
+      "grad_norm": 0.21176312397235778,
+      "learning_rate": 9.980684400076301e-06,
+      "loss": 0.5802,
+      "step": 90
+    },
+    {
+      "epoch": 0.08972146906581217,
+      "grad_norm": 0.1866670497645919,
+      "learning_rate": 9.980227543473243e-06,
+      "loss": 0.5983,
+      "step": 91
+    },
+    {
+      "epoch": 0.09070741927532659,
+      "grad_norm": 0.20887738100175152,
+      "learning_rate": 9.979765357703442e-06,
+      "loss": 0.5711,
+      "step": 92
+    },
+    {
+      "epoch": 0.09169336948484101,
+      "grad_norm": 0.18824727898046928,
+      "learning_rate": 9.97929784326147e-06,
+      "loss": 0.5842,
+      "step": 93
+    },
+    {
+      "epoch": 0.09267931969435543,
+      "grad_norm": 0.20872137142073366,
+      "learning_rate": 9.978825000647603e-06,
+      "loss": 0.5526,
+      "step": 94
+    },
+    {
+      "epoch": 0.09366526990386985,
+      "grad_norm": 0.20588660604802805,
+      "learning_rate": 9.978346830367804e-06,
+      "loss": 0.5772,
+      "step": 95
+    },
+    {
+      "epoch": 0.09465122011338427,
+      "grad_norm": 0.1897213779338987,
+      "learning_rate": 9.977863332933752e-06,
+      "loss": 0.589,
+      "step": 96
+    },
+    {
+      "epoch": 0.09563717032289869,
+      "grad_norm": 0.1938789139645962,
+      "learning_rate": 9.97737450886282e-06,
+      "loss": 0.5826,
+      "step": 97
+    },
+    {
+      "epoch": 0.0966231205324131,
+      "grad_norm": 0.18708886016881085,
+      "learning_rate": 9.976880358678083e-06,
+      "loss": 0.5747,
+      "step": 98
+    },
+    {
+      "epoch": 0.09760907074192754,
+      "grad_norm": 0.194768187909786,
+      "learning_rate": 9.97638088290831e-06,
+      "loss": 0.5743,
+      "step": 99
+    },
+    {
+      "epoch": 0.09859502095144196,
+      "grad_norm": 0.20855204340244599,
+      "learning_rate": 9.975876082087974e-06,
+      "loss": 0.5895,
+      "step": 100
+    },
+    {
+      "epoch": 0.09958097116095638,
+      "grad_norm": 0.19682699295375078,
+      "learning_rate": 9.975365956757245e-06,
+      "loss": 0.5815,
+      "step": 101
+    },
+    {
+      "epoch": 0.1005669213704708,
+      "grad_norm": 0.19761425189672296,
+      "learning_rate": 9.974850507461989e-06,
+      "loss": 0.57,
+      "step": 102
+    },
+    {
+      "epoch": 0.10155287157998522,
+      "grad_norm": 0.1969307392448049,
+      "learning_rate": 9.974329734753773e-06,
+      "loss": 0.6095,
+      "step": 103
+    },
+    {
+      "epoch": 0.10253882178949963,
+      "grad_norm": 0.19238160658361544,
+      "learning_rate": 9.973803639189857e-06,
+      "loss": 0.5953,
+      "step": 104
+    },
+    {
+      "epoch": 0.10352477199901405,
+      "grad_norm": 0.19536784964342985,
+      "learning_rate": 9.973272221333194e-06,
+      "loss": 0.5724,
+      "step": 105
+    },
+    {
+      "epoch": 0.10451072220852847,
+      "grad_norm": 0.18177641456857388,
+      "learning_rate": 9.972735481752438e-06,
+      "loss": 0.5994,
+      "step": 106
+    },
+    {
+      "epoch": 0.10549667241804289,
+      "grad_norm": 0.2058000188653258,
+      "learning_rate": 9.972193421021936e-06,
+      "loss": 0.5839,
+      "step": 107
+    },
+    {
+      "epoch": 0.10648262262755731,
+      "grad_norm": 0.20116963669280283,
+      "learning_rate": 9.971646039721727e-06,
+      "loss": 0.5845,
+      "step": 108
+    },
+    {
+      "epoch": 0.10746857283707173,
+      "grad_norm": 0.192644533678385,
+      "learning_rate": 9.971093338437545e-06,
+      "loss": 0.5943,
+      "step": 109
+    },
+    {
+      "epoch": 0.10845452304658615,
+      "grad_norm": 0.2051913783736844,
+      "learning_rate": 9.970535317760817e-06,
+      "loss": 0.5874,
+      "step": 110
+    },
+    {
+      "epoch": 0.10944047325610057,
+      "grad_norm": 0.18646892654806063,
+      "learning_rate": 9.96997197828866e-06,
+      "loss": 0.5855,
+      "step": 111
+    },
+    {
+      "epoch": 0.11042642346561499,
+      "grad_norm": 0.18808130693791184,
+      "learning_rate": 9.969403320623883e-06,
+      "loss": 0.5762,
+      "step": 112
+    },
+    {
+      "epoch": 0.1114123736751294,
+      "grad_norm": 0.18786557009902174,
+      "learning_rate": 9.968829345374988e-06,
+      "loss": 0.5782,
+      "step": 113
+    },
+    {
+      "epoch": 0.11239832388464382,
+      "grad_norm": 0.19710354354314336,
+      "learning_rate": 9.968250053156165e-06,
+      "loss": 0.5636,
+      "step": 114
+    },
+    {
+      "epoch": 0.11338427409415824,
+      "grad_norm": 0.19653609173201655,
+      "learning_rate": 9.967665444587298e-06,
+      "loss": 0.5575,
+      "step": 115
+    },
+    {
+      "epoch": 0.11437022430367266,
+      "grad_norm": 0.1934871588149603,
+      "learning_rate": 9.96707552029395e-06,
+      "loss": 0.5856,
+      "step": 116
+    },
+    {
+      "epoch": 0.11535617451318708,
+      "grad_norm": 0.21484146881712957,
+      "learning_rate": 9.966480280907383e-06,
+      "loss": 0.5825,
+      "step": 117
+    },
+    {
+      "epoch": 0.1163421247227015,
+      "grad_norm": 0.18883422828114707,
+      "learning_rate": 9.965879727064538e-06,
+      "loss": 0.5587,
+      "step": 118
+    },
+    {
+      "epoch": 0.11732807493221592,
+      "grad_norm": 0.18961757693952455,
+      "learning_rate": 9.965273859408052e-06,
+      "loss": 0.5752,
+      "step": 119
+    },
+    {
+      "epoch": 0.11831402514173034,
+      "grad_norm": 0.19297702343721243,
+      "learning_rate": 9.964662678586235e-06,
+      "loss": 0.5645,
+      "step": 120
+    },
+    {
+      "epoch": 0.11929997535124476,
+      "grad_norm": 0.2304035112033173,
+      "learning_rate": 9.964046185253098e-06,
+      "loss": 0.5786,
+      "step": 121
+    },
+    {
+      "epoch": 0.12028592556075918,
+      "grad_norm": 0.2506785823292142,
+      "learning_rate": 9.963424380068324e-06,
+      "loss": 0.583,
+      "step": 122
+    },
+    {
+      "epoch": 0.1212718757702736,
+      "grad_norm": 0.20430322582417668,
+      "learning_rate": 9.962797263697286e-06,
+      "loss": 0.596,
+      "step": 123
+    },
+    {
+      "epoch": 0.12225782597978802,
+      "grad_norm": 0.19713747960925884,
+      "learning_rate": 9.96216483681104e-06,
+      "loss": 0.5596,
+      "step": 124
+    },
+    {
+      "epoch": 0.12324377618930243,
+      "grad_norm": 0.195136749034251,
+      "learning_rate": 9.961527100086323e-06,
+      "loss": 0.5739,
+      "step": 125
+    },
+    {
+      "epoch": 0.12422972639881685,
+      "grad_norm": 0.18181708293146226,
+      "learning_rate": 9.960884054205556e-06,
+      "loss": 0.5757,
+      "step": 126
+    },
+    {
+      "epoch": 0.12521567660833127,
+      "grad_norm": 0.20850687238386326,
+      "learning_rate": 9.960235699856838e-06,
+      "loss": 0.5786,
+      "step": 127
+    },
+    {
+      "epoch": 0.1262016268178457,
+      "grad_norm": 0.6952482779983634,
+      "learning_rate": 9.959582037733952e-06,
+      "loss": 0.5802,
+      "step": 128
+    },
+    {
+      "epoch": 0.1271875770273601,
+      "grad_norm": 0.1950690815783342,
+      "learning_rate": 9.958923068536356e-06,
+      "loss": 0.6032,
+      "step": 129
+    },
+    {
+      "epoch": 0.12817352723687453,
+      "grad_norm": 0.19926639507294414,
+      "learning_rate": 9.958258792969195e-06,
+      "loss": 0.5768,
+      "step": 130
+    },
+    {
+      "epoch": 0.12915947744638895,
+      "grad_norm": 0.46729534031431685,
+      "learning_rate": 9.95758921174328e-06,
+      "loss": 0.5646,
+      "step": 131
+    },
+    {
+      "epoch": 0.13014542765590337,
+      "grad_norm": 0.21133226011241232,
+      "learning_rate": 9.956914325575114e-06,
+      "loss": 0.5527,
+      "step": 132
+    },
+    {
+      "epoch": 0.1311313778654178,
+      "grad_norm": 0.19752264024958172,
+      "learning_rate": 9.956234135186864e-06,
+      "loss": 0.5817,
+      "step": 133
+    },
+    {
+      "epoch": 0.1321173280749322,
+      "grad_norm": 0.43869306042549744,
+      "learning_rate": 9.955548641306379e-06,
+      "loss": 0.5625,
+      "step": 134
+    },
+    {
+      "epoch": 0.13310327828444662,
+      "grad_norm": 0.20043630296357293,
+      "learning_rate": 9.95485784466718e-06,
+      "loss": 0.5724,
+      "step": 135
+    },
+    {
+      "epoch": 0.13408922849396104,
+      "grad_norm": 0.19166523534205976,
+      "learning_rate": 9.954161746008468e-06,
+      "loss": 0.578,
+      "step": 136
+    },
+    {
+      "epoch": 0.13507517870347546,
+      "grad_norm": 0.1915364173625567,
+      "learning_rate": 9.953460346075112e-06,
+      "loss": 0.5667,
+      "step": 137
+    },
+    {
+      "epoch": 0.13606112891298988,
+      "grad_norm": 0.18716037179875467,
+      "learning_rate": 9.952753645617656e-06,
+      "loss": 0.5699,
+      "step": 138
+    },
+    {
+      "epoch": 0.1370470791225043,
+      "grad_norm": 0.19586717678531615,
+      "learning_rate": 9.952041645392313e-06,
+      "loss": 0.5787,
+      "step": 139
+    },
+    {
+      "epoch": 0.13803302933201872,
+      "grad_norm": 0.2144468317967703,
+      "learning_rate": 9.951324346160973e-06,
+      "loss": 0.5878,
+      "step": 140
+    },
+    {
+      "epoch": 0.13901897954153317,
+      "grad_norm": 0.24086052677356343,
+      "learning_rate": 9.95060174869119e-06,
+      "loss": 0.5489,
+      "step": 141
+    },
+    {
+      "epoch": 0.14000492975104759,
+      "grad_norm": 0.298794931994396,
+      "learning_rate": 9.94987385375619e-06,
+      "loss": 0.5677,
+      "step": 142
+    },
+    {
+      "epoch": 0.140990879960562,
+      "grad_norm": 0.1915597370513195,
+      "learning_rate": 9.949140662134873e-06,
+      "loss": 0.5816,
+      "step": 143
+    },
+    {
+      "epoch": 0.14197683017007642,
+      "grad_norm": 0.18685365109490332,
+      "learning_rate": 9.948402174611795e-06,
+      "loss": 0.5694,
+      "step": 144
+    },
+    {
+      "epoch": 0.14296278037959084,
+      "grad_norm": 0.19692674211623318,
+      "learning_rate": 9.947658391977188e-06,
+      "loss": 0.567,
+      "step": 145
+    },
+    {
+      "epoch": 0.14394873058910526,
+      "grad_norm": 0.1874193951834725,
+      "learning_rate": 9.94690931502695e-06,
+      "loss": 0.5965,
+      "step": 146
+    },
+    {
+      "epoch": 0.14493468079861968,
+      "grad_norm": 0.19259727731903561,
+      "learning_rate": 9.946154944562637e-06,
+      "loss": 0.5698,
+      "step": 147
+    },
+    {
+      "epoch": 0.1459206310081341,
+      "grad_norm": 0.9856315571854328,
+      "learning_rate": 9.945395281391478e-06,
+      "loss": 0.5816,
+      "step": 148
+    },
+    {
+      "epoch": 0.14690658121764852,
+      "grad_norm": 0.19217330617850037,
+      "learning_rate": 9.944630326326362e-06,
+      "loss": 0.5604,
+      "step": 149
+    },
+    {
+      "epoch": 0.14789253142716294,
+      "grad_norm": 0.18772651400583082,
+      "learning_rate": 9.94386008018584e-06,
+      "loss": 0.573,
+      "step": 150
+    },
+    {
+      "epoch": 0.14887848163667736,
+      "grad_norm": 0.5746457009545806,
+      "learning_rate": 9.943084543794123e-06,
+      "loss": 0.578,
+      "step": 151
+    },
+    {
+      "epoch": 0.14986443184619178,
+      "grad_norm": 0.20766753738887306,
+      "learning_rate": 9.94230371798109e-06,
+      "loss": 0.5667,
+      "step": 152
+    },
+    {
+      "epoch": 0.1508503820557062,
+      "grad_norm": 0.2004304147804513,
+      "learning_rate": 9.94151760358227e-06,
+      "loss": 0.5571,
+      "step": 153
+    },
+    {
+      "epoch": 0.1518363322652206,
+      "grad_norm": 0.20455822399303228,
+      "learning_rate": 9.940726201438862e-06,
+      "loss": 0.5672,
+      "step": 154
+    },
+    {
+      "epoch": 0.15282228247473503,
+      "grad_norm": 0.1838803288134749,
+      "learning_rate": 9.939929512397715e-06,
+      "loss": 0.5987,
+      "step": 155
+    },
+    {
+      "epoch": 0.15380823268424945,
+      "grad_norm": 0.20050980222504577,
+      "learning_rate": 9.939127537311337e-06,
+      "loss": 0.5699,
+      "step": 156
+    },
+    {
+      "epoch": 0.15479418289376387,
+      "grad_norm": 0.19441962825480735,
+      "learning_rate": 9.938320277037893e-06,
+      "loss": 0.5628,
+      "step": 157
+    },
+    {
+      "epoch": 0.1557801331032783,
+      "grad_norm": 0.19936545332740366,
+      "learning_rate": 9.937507732441206e-06,
+      "loss": 0.5605,
+      "step": 158
+    },
+    {
+      "epoch": 0.1567660833127927,
+      "grad_norm": 1.309212076590478,
+      "learning_rate": 9.93668990439075e-06,
+      "loss": 0.5693,
+      "step": 159
+    },
+    {
+      "epoch": 0.15775203352230713,
+      "grad_norm": 0.18988324352303748,
+      "learning_rate": 9.935866793761656e-06,
+      "loss": 0.5676,
+      "step": 160
+    },
+    {
+      "epoch": 0.15873798373182155,
+      "grad_norm": 0.19272518366856584,
+      "learning_rate": 9.935038401434702e-06,
+      "loss": 0.5371,
+      "step": 161
+    },
+    {
+      "epoch": 0.15972393394133597,
+      "grad_norm": 0.18550671681247388,
+      "learning_rate": 9.934204728296324e-06,
+      "loss": 0.5753,
+      "step": 162
+    },
+    {
+      "epoch": 0.16070988415085039,
+      "grad_norm": 0.19249215809324505,
+      "learning_rate": 9.933365775238609e-06,
+      "loss": 0.5609,
+      "step": 163
+    },
+    {
+      "epoch": 0.1616958343603648,
+      "grad_norm": 0.20411023032602532,
+      "learning_rate": 9.932521543159285e-06,
+      "loss": 0.5531,
+      "step": 164
+    },
+    {
+      "epoch": 0.16268178456987922,
+      "grad_norm": 0.19335472696117534,
+      "learning_rate": 9.931672032961742e-06,
+      "loss": 0.5469,
+      "step": 165
+    },
+    {
+      "epoch": 0.16366773477939364,
+      "grad_norm": 0.22420301199053938,
+      "learning_rate": 9.930817245555007e-06,
+      "loss": 0.5734,
+      "step": 166
+    },
+    {
+      "epoch": 0.16465368498890806,
+      "grad_norm": 0.2025719628644905,
+      "learning_rate": 9.929957181853758e-06,
+      "loss": 0.5653,
+      "step": 167
+    },
+    {
+      "epoch": 0.16563963519842248,
+      "grad_norm": 0.17704783553055983,
+      "learning_rate": 9.92909184277832e-06,
+      "loss": 0.5461,
+      "step": 168
+    },
+    {
+      "epoch": 0.1666255854079369,
+      "grad_norm": 0.18847786991320487,
+      "learning_rate": 9.928221229254661e-06,
+      "loss": 0.5734,
+      "step": 169
+    },
+    {
+      "epoch": 0.16761153561745132,
+      "grad_norm": 0.19669314062323434,
+      "learning_rate": 9.927345342214398e-06,
+      "loss": 0.5465,
+      "step": 170
+    },
+    {
+      "epoch": 0.16859748582696574,
+      "grad_norm": 0.19105027030988314,
+      "learning_rate": 9.926464182594781e-06,
+      "loss": 0.5818,
+      "step": 171
+    },
+    {
+      "epoch": 0.16958343603648016,
+      "grad_norm": 0.1874441302590153,
+      "learning_rate": 9.925577751338711e-06,
+      "loss": 0.5506,
+      "step": 172
+    },
+    {
+      "epoch": 0.17056938624599458,
+      "grad_norm": 0.17967781696372298,
+      "learning_rate": 9.924686049394728e-06,
+      "loss": 0.5244,
+      "step": 173
+    },
+    {
+      "epoch": 0.171555336455509,
+      "grad_norm": 0.19009914054945548,
+      "learning_rate": 9.923789077717007e-06,
+      "loss": 0.5577,
+      "step": 174
+    },
+    {
+      "epoch": 0.1725412866650234,
+      "grad_norm": 0.1931608565336059,
+      "learning_rate": 9.922886837265371e-06,
+      "loss": 0.5638,
+      "step": 175
+    },
+    {
+      "epoch": 0.17352723687453783,
+      "grad_norm": 0.26666031921399536,
+      "learning_rate": 9.921979329005271e-06,
+      "loss": 0.5409,
+      "step": 176
+    },
+    {
+      "epoch": 0.17451318708405225,
+      "grad_norm": 0.20832489811919228,
+      "learning_rate": 9.921066553907803e-06,
+      "loss": 0.5732,
+      "step": 177
+    },
+    {
+      "epoch": 0.17549913729356667,
+      "grad_norm": 0.18434435545608718,
+      "learning_rate": 9.920148512949697e-06,
+      "loss": 0.5578,
+      "step": 178
+    },
+    {
+      "epoch": 0.1764850875030811,
+      "grad_norm": 0.17758143999650278,
+      "learning_rate": 9.919225207113313e-06,
+      "loss": 0.5638,
+      "step": 179
+    },
+    {
+      "epoch": 0.1774710377125955,
+      "grad_norm": 0.1904346420359084,
+      "learning_rate": 9.918296637386648e-06,
+      "loss": 0.5635,
+      "step": 180
+    },
+    {
+      "epoch": 0.17845698792210993,
+      "grad_norm": 0.20043556886204436,
+      "learning_rate": 9.917362804763334e-06,
+      "loss": 0.5821,
+      "step": 181
+    },
+    {
+      "epoch": 0.17944293813162435,
+      "grad_norm": 0.3301171602767553,
+      "learning_rate": 9.91642371024263e-06,
+      "loss": 0.5734,
+      "step": 182
+    },
+    {
+      "epoch": 0.18042888834113877,
+      "grad_norm": 0.6608374464890221,
+      "learning_rate": 9.915479354829433e-06,
+      "loss": 0.5758,
+      "step": 183
+    },
+    {
+      "epoch": 0.18141483855065318,
+      "grad_norm": 0.18220942279149702,
+      "learning_rate": 9.91452973953426e-06,
+      "loss": 0.5357,
+      "step": 184
+    },
+    {
+      "epoch": 0.1824007887601676,
+      "grad_norm": 0.2203525160795268,
+      "learning_rate": 9.913574865373264e-06,
+      "loss": 0.5541,
+      "step": 185
+    },
+    {
+      "epoch": 0.18338673896968202,
+      "grad_norm": 0.18496823067115975,
+      "learning_rate": 9.912614733368218e-06,
+      "loss": 0.5693,
+      "step": 186
+    },
+    {
+      "epoch": 0.18437268917919644,
+      "grad_norm": 0.5019587689639134,
+      "learning_rate": 9.91164934454653e-06,
+      "loss": 0.5627,
+      "step": 187
+    },
+    {
+      "epoch": 0.18535863938871086,
+      "grad_norm": 0.2151624048884641,
+      "learning_rate": 9.910678699941227e-06,
+      "loss": 0.5584,
+      "step": 188
+    },
+    {
+      "epoch": 0.18634458959822528,
+      "grad_norm": 0.19157867884160143,
+      "learning_rate": 9.90970280059096e-06,
+      "loss": 0.5463,
+      "step": 189
+    },
+    {
+      "epoch": 0.1873305398077397,
+      "grad_norm": 0.19102624119101116,
+      "learning_rate": 9.90872164754001e-06,
+      "loss": 0.5479,
+      "step": 190
+    },
+    {
+      "epoch": 0.18831649001725412,
+      "grad_norm": 0.18358944790975948,
+      "learning_rate": 9.907735241838268e-06,
+      "loss": 0.5508,
+      "step": 191
+    },
+    {
+      "epoch": 0.18930244022676854,
+      "grad_norm": 0.17766251380399775,
+      "learning_rate": 9.906743584541256e-06,
+      "loss": 0.5637,
+      "step": 192
+    },
+    {
+      "epoch": 0.19028839043628296,
+      "grad_norm": 0.18203909053982936,
+      "learning_rate": 9.90574667671011e-06,
+      "loss": 0.5486,
+      "step": 193
+    },
+    {
+      "epoch": 0.19127434064579737,
+      "grad_norm": 0.18077750869311915,
+      "learning_rate": 9.904744519411588e-06,
+      "loss": 0.5719,
+      "step": 194
+    },
+    {
+      "epoch": 0.1922602908553118,
+      "grad_norm": 0.20529393815616548,
+      "learning_rate": 9.903737113718062e-06,
+      "loss": 0.5683,
+      "step": 195
+    },
+    {
+      "epoch": 0.1932462410648262,
+      "grad_norm": 0.17841515699100288,
+      "learning_rate": 9.90272446070752e-06,
+      "loss": 0.5439,
+      "step": 196
+    },
+    {
+      "epoch": 0.19423219127434063,
+      "grad_norm": 0.18054547942202176,
+      "learning_rate": 9.90170656146357e-06,
+      "loss": 0.5568,
+      "step": 197
+    },
+    {
+      "epoch": 0.19521814148385508,
+      "grad_norm": 0.32374590769366285,
+      "learning_rate": 9.900683417075427e-06,
+      "loss": 0.5561,
+      "step": 198
+    },
+    {
+      "epoch": 0.1962040916933695,
+      "grad_norm": 0.19010714932520278,
+      "learning_rate": 9.899655028637924e-06,
+      "loss": 0.5421,
+      "step": 199
+    },
+    {
+      "epoch": 0.19719004190288392,
+      "grad_norm": 0.1754484113927627,
+      "learning_rate": 9.898621397251503e-06,
+      "loss": 0.5575,
+      "step": 200
+    },
+    {
+      "epoch": 0.19817599211239834,
+      "grad_norm": 0.1858887349509897,
+      "learning_rate": 9.897582524022216e-06,
+      "loss": 0.5396,
+      "step": 201
+    },
+    {
+      "epoch": 0.19916194232191275,
+      "grad_norm": 0.1737758449718213,
+      "learning_rate": 9.896538410061724e-06,
+      "loss": 0.5496,
+      "step": 202
+    },
+    {
+      "epoch": 0.20014789253142717,
+      "grad_norm": 0.19356547299150076,
+      "learning_rate": 9.895489056487298e-06,
+      "loss": 0.5549,
+      "step": 203
+    },
+    {
+      "epoch": 0.2011338427409416,
+      "grad_norm": 0.1776417983268236,
+      "learning_rate": 9.894434464421817e-06,
+      "loss": 0.5696,
+      "step": 204
+    },
+    {
+      "epoch": 0.202119792950456,
+      "grad_norm": 0.18109120205294296,
+      "learning_rate": 9.893374634993756e-06,
+      "loss": 0.5504,
+      "step": 205
+    },
+    {
+      "epoch": 0.20310574315997043,
+      "grad_norm": 0.18540548568684925,
+      "learning_rate": 9.892309569337208e-06,
+      "loss": 0.5532,
+      "step": 206
+    },
+    {
+      "epoch": 0.20409169336948485,
+      "grad_norm": 0.17913710095312485,
+      "learning_rate": 9.891239268591858e-06,
+      "loss": 0.5504,
+      "step": 207
+    },
+    {
+      "epoch": 0.20507764357899927,
+      "grad_norm": 0.18027765377069868,
+      "learning_rate": 9.890163733903003e-06,
+      "loss": 0.5569,
+      "step": 208
+    },
+    {
+      "epoch": 0.2060635937885137,
+      "grad_norm": 0.18205724705660326,
+      "learning_rate": 9.889082966421529e-06,
+      "loss": 0.5585,
+      "step": 209
+    },
+    {
+      "epoch": 0.2070495439980281,
+      "grad_norm": 0.1892943540336806,
+      "learning_rate": 9.887996967303928e-06,
+      "loss": 0.561,
+      "step": 210
+    },
+    {
+      "epoch": 0.20803549420754253,
+      "grad_norm": 0.352372345851401,
+      "learning_rate": 9.88690573771229e-06,
+      "loss": 0.5594,
+      "step": 211
+    },
+    {
+      "epoch": 0.20902144441705695,
+      "grad_norm": 0.17451224769434992,
+      "learning_rate": 9.885809278814307e-06,
+      "loss": 0.5637,
+      "step": 212
+    },
+    {
+      "epoch": 0.21000739462657136,
+      "grad_norm": 0.19118016535636867,
+      "learning_rate": 9.884707591783253e-06,
+      "loss": 0.5532,
+      "step": 213
+    },
+    {
+      "epoch": 0.21099334483608578,
+      "grad_norm": 0.17955764775368546,
+      "learning_rate": 9.88360067779801e-06,
+      "loss": 0.5423,
+      "step": 214
+    },
+    {
+      "epoch": 0.2119792950456002,
+      "grad_norm": 0.19225335942620098,
+      "learning_rate": 9.882488538043044e-06,
+      "loss": 0.5498,
+      "step": 215
+    },
+    {
+      "epoch": 0.21296524525511462,
+      "grad_norm": 0.17835521350051764,
+      "learning_rate": 9.881371173708421e-06,
+      "loss": 0.565,
+      "step": 216
+    },
+    {
+      "epoch": 0.21395119546462904,
+      "grad_norm": 0.17648927779006615,
+      "learning_rate": 9.88024858598979e-06,
+      "loss": 0.5699,
+      "step": 217
+    },
+    {
+      "epoch": 0.21493714567414346,
+      "grad_norm": 0.23186770896350944,
+      "learning_rate": 9.879120776088396e-06,
+      "loss": 0.5336,
+      "step": 218
+    },
+    {
+      "epoch": 0.21592309588365788,
+      "grad_norm": 0.1865552812650371,
+      "learning_rate": 9.877987745211065e-06,
+      "loss": 0.5603,
+      "step": 219
+    },
+    {
+      "epoch": 0.2169090460931723,
+      "grad_norm": 0.187354863042195,
+      "learning_rate": 9.876849494570216e-06,
+      "loss": 0.5811,
+      "step": 220
+    },
+    {
+      "epoch": 0.21789499630268672,
+      "grad_norm": 0.17681306208150294,
+      "learning_rate": 9.87570602538385e-06,
+      "loss": 0.5486,
+      "step": 221
+    },
+    {
+      "epoch": 0.21888094651220114,
+      "grad_norm": 0.19341784130459566,
+      "learning_rate": 9.874557338875554e-06,
+      "loss": 0.5473,
+      "step": 222
+    },
+    {
+      "epoch": 0.21986689672171555,
+      "grad_norm": 0.18861128025094429,
+      "learning_rate": 9.873403436274495e-06,
+      "loss": 0.5463,
+      "step": 223
+    },
+    {
+      "epoch": 0.22085284693122997,
+      "grad_norm": 0.18384453190517155,
+      "learning_rate": 9.872244318815428e-06,
+      "loss": 0.5435,
+      "step": 224
+    },
+    {
+      "epoch": 0.2218387971407444,
+      "grad_norm": 0.18217399013944716,
+      "learning_rate": 9.871079987738681e-06,
+      "loss": 0.5576,
+      "step": 225
+    },
+    {
+      "epoch": 0.2228247473502588,
+      "grad_norm": 0.18430432846747724,
+      "learning_rate": 9.869910444290162e-06,
+      "loss": 0.5578,
+      "step": 226
+    },
+    {
+      "epoch": 0.22381069755977323,
+      "grad_norm": 0.18089824279506567,
+      "learning_rate": 9.868735689721363e-06,
+      "loss": 0.5618,
+      "step": 227
+    },
+    {
+      "epoch": 0.22479664776928765,
+      "grad_norm": 0.17403827133499156,
+      "learning_rate": 9.867555725289344e-06,
+      "loss": 0.5654,
+      "step": 228
+    },
+    {
+      "epoch": 0.22578259797880207,
+      "grad_norm": 0.18781609637012886,
+      "learning_rate": 9.866370552256746e-06,
+      "loss": 0.5405,
+      "step": 229
+    },
+    {
+      "epoch": 0.2267685481883165,
+      "grad_norm": 0.1762793424800406,
+      "learning_rate": 9.865180171891778e-06,
+      "loss": 0.5504,
+      "step": 230
+    },
+    {
+      "epoch": 0.2277544983978309,
+      "grad_norm": 0.18853685496248657,
+      "learning_rate": 9.863984585468226e-06,
+      "loss": 0.5503,
+      "step": 231
+    },
+    {
+      "epoch": 0.22874044860734533,
+      "grad_norm": 0.1823533722312669,
+      "learning_rate": 9.862783794265448e-06,
+      "loss": 0.5607,
+      "step": 232
+    },
+    {
+      "epoch": 0.22972639881685974,
+      "grad_norm": 0.18201466636582803,
+      "learning_rate": 9.861577799568364e-06,
+      "loss": 0.5387,
+      "step": 233
+    },
+    {
+      "epoch": 0.23071234902637416,
+      "grad_norm": 0.1862180638822258,
+      "learning_rate": 9.860366602667469e-06,
+      "loss": 0.5594,
+      "step": 234
+    },
+    {
+      "epoch": 0.23169829923588858,
+      "grad_norm": 0.1755168236218862,
+      "learning_rate": 9.85915020485882e-06,
+      "loss": 0.53,
+      "step": 235
+    },
+    {
+      "epoch": 0.232684249445403,
+      "grad_norm": 0.1824850169156925,
+      "learning_rate": 9.857928607444045e-06,
+      "loss": 0.5385,
+      "step": 236
+    },
+    {
+      "epoch": 0.23367019965491742,
+      "grad_norm": 0.18537873529999935,
+      "learning_rate": 9.85670181173033e-06,
+      "loss": 0.5407,
+      "step": 237
+    },
+    {
+      "epoch": 0.23465614986443184,
+      "grad_norm": 0.20704735012455386,
+      "learning_rate": 9.855469819030425e-06,
+      "loss": 0.558,
+      "step": 238
+    },
+    {
+      "epoch": 0.23564210007394626,
+      "grad_norm": 0.18429423614956533,
+      "learning_rate": 9.854232630662647e-06,
+      "loss": 0.5399,
+      "step": 239
+    },
+    {
+      "epoch": 0.23662805028346068,
+      "grad_norm": 0.19422829306134085,
+      "learning_rate": 9.852990247950863e-06,
+      "loss": 0.5682,
+      "step": 240
+    },
+    {
+      "epoch": 0.2376140004929751,
+      "grad_norm": 0.1913164213893665,
+      "learning_rate": 9.851742672224506e-06,
+      "loss": 0.5513,
+      "step": 241
+    },
+    {
+      "epoch": 0.23859995070248952,
+      "grad_norm": 0.1788406214546875,
+      "learning_rate": 9.850489904818561e-06,
+      "loss": 0.5419,
+      "step": 242
+    },
+    {
+      "epoch": 0.23958590091200394,
+      "grad_norm": 0.18794434492520137,
+      "learning_rate": 9.849231947073571e-06,
+      "loss": 0.5604,
+      "step": 243
+    },
+    {
+      "epoch": 0.24057185112151835,
+      "grad_norm": 0.18685472437860745,
+      "learning_rate": 9.847968800335635e-06,
+      "loss": 0.5419,
+      "step": 244
+    },
+    {
+      "epoch": 0.24155780133103277,
+      "grad_norm": 0.18174774064309807,
+      "learning_rate": 9.846700465956399e-06,
+      "loss": 0.562,
+      "step": 245
+    },
+    {
+      "epoch": 0.2425437515405472,
+      "grad_norm": 0.1824025047659135,
+      "learning_rate": 9.845426945293064e-06,
+      "loss": 0.5365,
+      "step": 246
+    },
+    {
+      "epoch": 0.2435297017500616,
+      "grad_norm": 0.23426015842358192,
+      "learning_rate": 9.84414823970838e-06,
+      "loss": 0.5302,
+      "step": 247
+    },
+    {
+      "epoch": 0.24451565195957603,
+      "grad_norm": 0.1931454250333834,
+      "learning_rate": 9.842864350570645e-06,
+      "loss": 0.5161,
+      "step": 248
+    },
+    {
+      "epoch": 0.24550160216909045,
+      "grad_norm": 0.18283942948337056,
+      "learning_rate": 9.8415752792537e-06,
+      "loss": 0.5548,
+      "step": 249
+    },
+    {
+      "epoch": 0.24648755237860487,
+      "grad_norm": 0.19154911738206512,
+      "learning_rate": 9.840281027136943e-06,
+      "loss": 0.5597,
+      "step": 250
+    },
+    {
+      "epoch": 0.2474735025881193,
+      "grad_norm": 0.18088172165095986,
+      "learning_rate": 9.838981595605301e-06,
+      "loss": 0.5592,
+      "step": 251
+    },
+    {
+      "epoch": 0.2484594527976337,
+      "grad_norm": 0.19847882231986305,
+      "learning_rate": 9.837676986049253e-06,
+      "loss": 0.5424,
+      "step": 252
+    },
+    {
+      "epoch": 0.24944540300714813,
+      "grad_norm": 0.190254367466553,
+      "learning_rate": 9.836367199864814e-06,
+      "loss": 0.5675,
+      "step": 253
+    },
+    {
+      "epoch": 0.25043135321666254,
+      "grad_norm": 0.1863957829966417,
+      "learning_rate": 9.835052238453543e-06,
+      "loss": 0.5467,
+      "step": 254
+    },
+    {
+      "epoch": 0.25141730342617696,
+      "grad_norm": 0.1895012556382713,
+      "learning_rate": 9.833732103222531e-06,
+      "loss": 0.5455,
+      "step": 255
+    },
+    {
+      "epoch": 0.2524032536356914,
+      "grad_norm": 0.18702548146488163,
+      "learning_rate": 9.832406795584412e-06,
+      "loss": 0.5611,
+      "step": 256
+    },
+    {
+      "epoch": 0.2533892038452058,
+      "grad_norm": 0.1918046148589585,
+      "learning_rate": 9.831076316957348e-06,
+      "loss": 0.5472,
+      "step": 257
+    },
+    {
+      "epoch": 0.2543751540547202,
+      "grad_norm": 0.19638448364666938,
+      "learning_rate": 9.829740668765037e-06,
+      "loss": 0.5548,
+      "step": 258
+    },
+    {
+      "epoch": 0.25536110426423464,
+      "grad_norm": 0.18213286342958498,
+      "learning_rate": 9.828399852436714e-06,
+      "loss": 0.5433,
+      "step": 259
+    },
+    {
+      "epoch": 0.25634705447374906,
+      "grad_norm": 0.20040305338002562,
+      "learning_rate": 9.827053869407134e-06,
+      "loss": 0.5809,
+      "step": 260
+    },
+    {
+      "epoch": 0.2573330046832635,
+      "grad_norm": 0.17823037993061608,
+      "learning_rate": 9.825702721116587e-06,
+      "loss": 0.5659,
+      "step": 261
+    },
+    {
+      "epoch": 0.2583189548927779,
+      "grad_norm": 0.1906176349922938,
+      "learning_rate": 9.824346409010895e-06,
+      "loss": 0.5542,
+      "step": 262
+    },
+    {
+      "epoch": 0.2593049051022923,
+      "grad_norm": 0.18236764853176807,
+      "learning_rate": 9.822984934541393e-06,
+      "loss": 0.5422,
+      "step": 263
+    },
+    {
+      "epoch": 0.26029085531180673,
+      "grad_norm": 0.3446126733000217,
+      "learning_rate": 9.821618299164953e-06,
+      "loss": 0.5439,
+      "step": 264
+    },
+    {
+      "epoch": 0.26127680552132115,
+      "grad_norm": 0.20311944757484177,
+      "learning_rate": 9.820246504343958e-06,
+      "loss": 0.5606,
+      "step": 265
+    },
+    {
+      "epoch": 0.2622627557308356,
+      "grad_norm": 0.1799513865731101,
+      "learning_rate": 9.818869551546319e-06,
+      "loss": 0.5256,
+      "step": 266
+    },
+    {
+      "epoch": 0.26324870594035,
+      "grad_norm": 0.18272493636473558,
+      "learning_rate": 9.817487442245468e-06,
+      "loss": 0.576,
+      "step": 267
+    },
+    {
+      "epoch": 0.2642346561498644,
+      "grad_norm": 0.21962884787364956,
+      "learning_rate": 9.816100177920349e-06,
+      "loss": 0.5376,
+      "step": 268
+    },
+    {
+      "epoch": 0.26522060635937883,
+      "grad_norm": 0.19015374963011628,
+      "learning_rate": 9.814707760055427e-06,
+      "loss": 0.5398,
+      "step": 269
+    },
+    {
+      "epoch": 0.26620655656889325,
+      "grad_norm": 0.20512522933688232,
+      "learning_rate": 9.813310190140676e-06,
+      "loss": 0.5708,
+      "step": 270
+    },
+    {
+      "epoch": 0.26719250677840767,
+      "grad_norm": 0.1807567800669277,
+      "learning_rate": 9.81190746967159e-06,
+      "loss": 0.5646,
+      "step": 271
+    },
+    {
+      "epoch": 0.2681784569879221,
+      "grad_norm": 0.2891229320005016,
+      "learning_rate": 9.810499600149166e-06,
+      "loss": 0.5372,
+      "step": 272
+    },
+    {
+      "epoch": 0.2691644071974365,
+      "grad_norm": 0.1806821909652169,
+      "learning_rate": 9.809086583079923e-06,
+      "loss": 0.5237,
+      "step": 273
+    },
+    {
+      "epoch": 0.2701503574069509,
+      "grad_norm": 0.18982968914898915,
+      "learning_rate": 9.807668419975876e-06,
+      "loss": 0.5571,
+      "step": 274
+    },
+    {
+      "epoch": 0.27113630761646534,
+      "grad_norm": 0.20260892553853443,
+      "learning_rate": 9.806245112354552e-06,
+      "loss": 0.5324,
+      "step": 275
+    },
+    {
+      "epoch": 0.27212225782597976,
+      "grad_norm": 0.18353665591257465,
+      "learning_rate": 9.804816661738984e-06,
+      "loss": 0.5298,
+      "step": 276
+    },
+    {
+      "epoch": 0.2731082080354942,
+      "grad_norm": 0.1776571164680753,
+      "learning_rate": 9.803383069657706e-06,
+      "loss": 0.5418,
+      "step": 277
+    },
+    {
+      "epoch": 0.2740941582450086,
+      "grad_norm": 0.1870562181329872,
+      "learning_rate": 9.801944337644755e-06,
+      "loss": 0.5607,
+      "step": 278
+    },
+    {
+      "epoch": 0.275080108454523,
+      "grad_norm": 0.20423434175503252,
+      "learning_rate": 9.800500467239666e-06,
+      "loss": 0.5466,
+      "step": 279
+    },
+    {
+      "epoch": 0.27606605866403744,
+      "grad_norm": 1.051465429104708,
+      "learning_rate": 9.799051459987478e-06,
+      "loss": 0.556,
+      "step": 280
+    },
+    {
+      "epoch": 0.27705200887355186,
+      "grad_norm": 0.21609582147081427,
+      "learning_rate": 9.797597317438719e-06,
+      "loss": 0.5555,
+      "step": 281
+    },
+    {
+      "epoch": 0.27803795908306633,
+      "grad_norm": 0.1876184929215964,
+      "learning_rate": 9.796138041149416e-06,
+      "loss": 0.5652,
+      "step": 282
+    },
+    {
+      "epoch": 0.27902390929258075,
+      "grad_norm": 0.18576235687638093,
+      "learning_rate": 9.794673632681093e-06,
+      "loss": 0.5323,
+      "step": 283
+    },
+    {
+      "epoch": 0.28000985950209517,
+      "grad_norm": 0.18728666616264814,
+      "learning_rate": 9.793204093600758e-06,
+      "loss": 0.5513,
+      "step": 284
+    },
+    {
+      "epoch": 0.2809958097116096,
+      "grad_norm": 0.34552956735847234,
+      "learning_rate": 9.791729425480917e-06,
+      "loss": 0.5436,
+      "step": 285
+    },
+    {
+      "epoch": 0.281981759921124,
+      "grad_norm": 0.1892052352760262,
+      "learning_rate": 9.790249629899555e-06,
+      "loss": 0.5698,
+      "step": 286
+    },
+    {
+      "epoch": 0.28296771013063843,
+      "grad_norm": 0.18248732304052728,
+      "learning_rate": 9.788764708440154e-06,
+      "loss": 0.5355,
+      "step": 287
+    },
+    {
+      "epoch": 0.28395366034015285,
+      "grad_norm": 0.1985966071161362,
+      "learning_rate": 9.787274662691677e-06,
+      "loss": 0.5446,
+      "step": 288
+    },
+    {
+      "epoch": 0.28493961054966727,
+      "grad_norm": 0.18044086283569283,
+      "learning_rate": 9.785779494248566e-06,
+      "loss": 0.5313,
+      "step": 289
+    },
+    {
+      "epoch": 0.2859255607591817,
+      "grad_norm": 0.1827212667367703,
+      "learning_rate": 9.784279204710751e-06,
+      "loss": 0.5566,
+      "step": 290
+    },
+    {
+      "epoch": 0.2869115109686961,
+      "grad_norm": 0.20487472968156348,
+      "learning_rate": 9.782773795683638e-06,
+      "loss": 0.5467,
+      "step": 291
+    },
+    {
+      "epoch": 0.2878974611782105,
+      "grad_norm": 0.17915713143387146,
+      "learning_rate": 9.781263268778112e-06,
+      "loss": 0.5555,
+      "step": 292
+    },
+    {
+      "epoch": 0.28888341138772494,
+      "grad_norm": 0.28311912335362205,
+      "learning_rate": 9.779747625610536e-06,
+      "loss": 0.5331,
+      "step": 293
+    },
+    {
+      "epoch": 0.28986936159723936,
+      "grad_norm": 0.19046021912368227,
+      "learning_rate": 9.778226867802748e-06,
+      "loss": 0.5458,
+      "step": 294
+    },
+    {
+      "epoch": 0.2908553118067538,
+      "grad_norm": 0.1949435939962373,
+      "learning_rate": 9.776700996982054e-06,
+      "loss": 0.5417,
+      "step": 295
+    },
+    {
+      "epoch": 0.2918412620162682,
+      "grad_norm": 0.17508894414087878,
+      "learning_rate": 9.775170014781235e-06,
+      "loss": 0.5303,
+      "step": 296
+    },
+    {
+      "epoch": 0.2928272122257826,
+      "grad_norm": 0.186462145746406,
+      "learning_rate": 9.773633922838545e-06,
+      "loss": 0.5335,
+      "step": 297
+    },
+    {
+      "epoch": 0.29381316243529704,
+      "grad_norm": 0.18907878327589916,
+      "learning_rate": 9.772092722797699e-06,
+      "loss": 0.5417,
+      "step": 298
+    },
+    {
+      "epoch": 0.29479911264481146,
+      "grad_norm": 0.18392737568551656,
+      "learning_rate": 9.770546416307883e-06,
+      "loss": 0.5471,
+      "step": 299
+    },
+    {
+      "epoch": 0.2957850628543259,
+      "grad_norm": 0.19021871221942088,
+      "learning_rate": 9.768995005023743e-06,
+      "loss": 0.5446,
+      "step": 300
+    },
+    {
+      "epoch": 0.2967710130638403,
+      "grad_norm": 0.18497373304429987,
+      "learning_rate": 9.76743849060539e-06,
+      "loss": 0.5287,
+      "step": 301
+    },
+    {
+      "epoch": 0.2977569632733547,
+      "grad_norm": 0.18659210300540374,
+      "learning_rate": 9.765876874718399e-06,
+      "loss": 0.5639,
+      "step": 302
+    },
+    {
+      "epoch": 0.29874291348286913,
+      "grad_norm": 0.17894062952145798,
+      "learning_rate": 9.764310159033797e-06,
+      "loss": 0.5553,
+      "step": 303
+    },
+    {
+      "epoch": 0.29972886369238355,
+      "grad_norm": 0.19551203985949925,
+      "learning_rate": 9.76273834522807e-06,
+      "loss": 0.5806,
+      "step": 304
+    },
+    {
+      "epoch": 0.30071481390189797,
+      "grad_norm": 0.18731748927278455,
+      "learning_rate": 9.761161434983166e-06,
+      "loss": 0.5663,
+      "step": 305
+    },
+    {
+      "epoch": 0.3017007641114124,
+      "grad_norm": 0.17968426298994367,
+      "learning_rate": 9.759579429986479e-06,
+      "loss": 0.5437,
+      "step": 306
+    },
+    {
+      "epoch": 0.3026867143209268,
+      "grad_norm": 0.22204292376132498,
+      "learning_rate": 9.757992331930855e-06,
+      "loss": 0.5432,
+      "step": 307
+    },
+    {
+      "epoch": 0.3036726645304412,
+      "grad_norm": 0.18184753314796798,
+      "learning_rate": 9.756400142514593e-06,
+      "loss": 0.548,
+      "step": 308
+    },
+    {
+      "epoch": 0.30465861473995565,
+      "grad_norm": 0.20647974009986644,
+      "learning_rate": 9.754802863441441e-06,
+      "loss": 0.5623,
+      "step": 309
+    },
+    {
+      "epoch": 0.30564456494947007,
+      "grad_norm": 0.18725507073578004,
+      "learning_rate": 9.75320049642059e-06,
+      "loss": 0.5568,
+      "step": 310
+    },
+    {
+      "epoch": 0.3066305151589845,
+      "grad_norm": 0.1800455366542249,
+      "learning_rate": 9.751593043166673e-06,
+      "loss": 0.5331,
+      "step": 311
+    },
+    {
+      "epoch": 0.3076164653684989,
+      "grad_norm": 0.1841426520536834,
+      "learning_rate": 9.749980505399777e-06,
+      "loss": 0.5407,
+      "step": 312
+    },
+    {
+      "epoch": 0.3086024155780133,
+      "grad_norm": 0.19659367711913434,
+      "learning_rate": 9.748362884845417e-06,
+      "loss": 0.5686,
+      "step": 313
+    },
+    {
+      "epoch": 0.30958836578752774,
+      "grad_norm": 0.17269369509189947,
+      "learning_rate": 9.74674018323455e-06,
+      "loss": 0.5309,
+      "step": 314
+    },
+    {
+      "epoch": 0.31057431599704216,
+      "grad_norm": 0.18860615825895338,
+      "learning_rate": 9.745112402303577e-06,
+      "loss": 0.5358,
+      "step": 315
+    },
+    {
+      "epoch": 0.3115602662065566,
+      "grad_norm": 0.18729708383811847,
+      "learning_rate": 9.74347954379433e-06,
+      "loss": 0.5558,
+      "step": 316
+    },
+    {
+      "epoch": 0.312546216416071,
+      "grad_norm": 0.1763617382688214,
+      "learning_rate": 9.741841609454067e-06,
+      "loss": 0.5358,
+      "step": 317
+    },
+    {
+      "epoch": 0.3135321666255854,
+      "grad_norm": 0.17979141365552667,
+      "learning_rate": 9.740198601035489e-06,
+      "loss": 0.5501,
+      "step": 318
+    },
+    {
+      "epoch": 0.31451811683509984,
+      "grad_norm": 0.1803838126614993,
+      "learning_rate": 9.738550520296722e-06,
+      "loss": 0.5325,
+      "step": 319
+    },
+    {
+      "epoch": 0.31550406704461426,
+      "grad_norm": 0.18884339503707848,
+      "learning_rate": 9.736897369001315e-06,
+      "loss": 0.5674,
+      "step": 320
+    },
+    {
+      "epoch": 0.3164900172541287,
+      "grad_norm": 0.1865791726288368,
+      "learning_rate": 9.735239148918251e-06,
+      "loss": 0.5402,
+      "step": 321
+    },
+    {
+      "epoch": 0.3174759674636431,
+      "grad_norm": 0.18694529564515183,
+      "learning_rate": 9.733575861821934e-06,
+      "loss": 0.5589,
+      "step": 322
+    },
+    {
+      "epoch": 0.3184619176731575,
+      "grad_norm": 0.20535318492014717,
+      "learning_rate": 9.731907509492185e-06,
+      "loss": 0.5424,
+      "step": 323
+    },
+    {
+      "epoch": 0.31944786788267193,
+      "grad_norm": 0.2071518418750301,
+      "learning_rate": 9.730234093714253e-06,
+      "loss": 0.5575,
+      "step": 324
+    },
+    {
+      "epoch": 0.32043381809218635,
+      "grad_norm": 0.17481253054592302,
+      "learning_rate": 9.7285556162788e-06,
+      "loss": 0.5409,
+      "step": 325
+    },
+    {
+      "epoch": 0.32141976830170077,
+      "grad_norm": 0.18594393577387902,
+      "learning_rate": 9.726872078981906e-06,
+      "loss": 0.5412,
+      "step": 326
+    },
+    {
+      "epoch": 0.3224057185112152,
+      "grad_norm": 0.19164197402550306,
+      "learning_rate": 9.725183483625065e-06,
+      "loss": 0.5555,
+      "step": 327
+    },
+    {
+      "epoch": 0.3233916687207296,
+      "grad_norm": 0.186053937689582,
+      "learning_rate": 9.723489832015183e-06,
+      "loss": 0.5718,
+      "step": 328
+    },
+    {
+      "epoch": 0.324377618930244,
+      "grad_norm": 0.1875853203551617,
+      "learning_rate": 9.721791125964578e-06,
+      "loss": 0.5575,
+      "step": 329
+    },
+    {
+      "epoch": 0.32536356913975845,
+      "grad_norm": 0.18819870895985158,
+      "learning_rate": 9.720087367290977e-06,
+      "loss": 0.5328,
+      "step": 330
+    },
+    {
+      "epoch": 0.32634951934927287,
+      "grad_norm": 0.18068715131695356,
+      "learning_rate": 9.71837855781751e-06,
+      "loss": 0.5326,
+      "step": 331
+    },
+    {
+      "epoch": 0.3273354695587873,
+      "grad_norm": 0.18016311746756167,
+      "learning_rate": 9.716664699372715e-06,
+      "loss": 0.533,
+      "step": 332
+    },
+    {
+      "epoch": 0.3283214197683017,
+      "grad_norm": 0.18207253599137066,
+      "learning_rate": 9.714945793790534e-06,
+      "loss": 0.5226,
+      "step": 333
+    },
+    {
+      "epoch": 0.3293073699778161,
+      "grad_norm": 0.20785211521181185,
+      "learning_rate": 9.713221842910304e-06,
+      "loss": 0.5423,
+      "step": 334
+    },
+    {
+      "epoch": 0.33029332018733054,
+      "grad_norm": 0.1845681139457901,
+      "learning_rate": 9.711492848576765e-06,
+      "loss": 0.536,
+      "step": 335
+    },
+    {
+      "epoch": 0.33127927039684496,
+      "grad_norm": 0.21007691072498666,
+      "learning_rate": 9.709758812640054e-06,
+      "loss": 0.5516,
+      "step": 336
+    },
+    {
+      "epoch": 0.3322652206063594,
+      "grad_norm": 0.1941130143828044,
+      "learning_rate": 9.708019736955701e-06,
+      "loss": 0.5592,
+      "step": 337
+    },
+    {
+      "epoch": 0.3332511708158738,
+      "grad_norm": 0.18770453694072897,
+      "learning_rate": 9.706275623384633e-06,
+      "loss": 0.5508,
+      "step": 338
+    },
+    {
+      "epoch": 0.3342371210253882,
+      "grad_norm": 0.18903620940691857,
+      "learning_rate": 9.70452647379316e-06,
+      "loss": 0.5311,
+      "step": 339
+    },
+    {
+      "epoch": 0.33522307123490264,
+      "grad_norm": 0.18273765065473813,
+      "learning_rate": 9.702772290052992e-06,
+      "loss": 0.5313,
+      "step": 340
+    },
+    {
+      "epoch": 0.33620902144441706,
+      "grad_norm": 0.1942003299384735,
+      "learning_rate": 9.701013074041213e-06,
+      "loss": 0.5437,
+      "step": 341
+    },
+    {
+      "epoch": 0.3371949716539315,
+      "grad_norm": 0.3108557907368959,
+      "learning_rate": 9.699248827640302e-06,
+      "loss": 0.5572,
+      "step": 342
+    },
+    {
+      "epoch": 0.3381809218634459,
+      "grad_norm": 0.1782316909211875,
+      "learning_rate": 9.697479552738117e-06,
+      "loss": 0.5283,
+      "step": 343
+    },
+    {
+      "epoch": 0.3391668720729603,
+      "grad_norm": 0.18682895440603317,
+      "learning_rate": 9.695705251227893e-06,
+      "loss": 0.5473,
+      "step": 344
+    },
+    {
+      "epoch": 0.34015282228247473,
+      "grad_norm": 0.18299858006665634,
+      "learning_rate": 9.693925925008251e-06,
+      "loss": 0.5547,
+      "step": 345
+    },
+    {
+      "epoch": 0.34113877249198915,
+      "grad_norm": 0.19762058128533963,
+      "learning_rate": 9.692141575983189e-06,
+      "loss": 0.5417,
+      "step": 346
+    },
+    {
+      "epoch": 0.34212472270150357,
+      "grad_norm": 0.2009208919552246,
+      "learning_rate": 9.69035220606207e-06,
+      "loss": 0.5515,
+      "step": 347
+    },
+    {
+      "epoch": 0.343110672911018,
+      "grad_norm": 0.17113169383263752,
+      "learning_rate": 9.68855781715964e-06,
+      "loss": 0.5305,
+      "step": 348
+    },
+    {
+      "epoch": 0.3440966231205324,
+      "grad_norm": 0.18805578103017412,
+      "learning_rate": 9.686758411196009e-06,
+      "loss": 0.5025,
+      "step": 349
+    },
+    {
+      "epoch": 0.3450825733300468,
+      "grad_norm": 0.21176364012019433,
+      "learning_rate": 9.68495399009666e-06,
+      "loss": 0.5279,
+      "step": 350
+    },
+    {
+      "epoch": 0.34606852353956125,
+      "grad_norm": 0.18507930814721,
+      "learning_rate": 9.683144555792441e-06,
+      "loss": 0.5272,
+      "step": 351
+    },
+    {
+      "epoch": 0.34705447374907566,
+      "grad_norm": 0.18608103510728297,
+      "learning_rate": 9.681330110219563e-06,
+      "loss": 0.5352,
+      "step": 352
+    },
+    {
+      "epoch": 0.3480404239585901,
+      "grad_norm": 0.1768125300825402,
+      "learning_rate": 9.6795106553196e-06,
+      "loss": 0.5548,
+      "step": 353
+    },
+    {
+      "epoch": 0.3490263741681045,
+      "grad_norm": 0.19221618896369344,
+      "learning_rate": 9.677686193039489e-06,
+      "loss": 0.5351,
+      "step": 354
+    },
+    {
+      "epoch": 0.3500123243776189,
+      "grad_norm": 0.17206225544086962,
+      "learning_rate": 9.67585672533152e-06,
+      "loss": 0.5293,
+      "step": 355
+    },
+    {
+      "epoch": 0.35099827458713334,
+      "grad_norm": 0.17510162004878352,
+      "learning_rate": 9.674022254153345e-06,
+      "loss": 0.561,
+      "step": 356
+    },
+    {
+      "epoch": 0.35198422479664776,
+      "grad_norm": 0.18027683086974733,
+      "learning_rate": 9.672182781467967e-06,
+      "loss": 0.5412,
+      "step": 357
+    },
+    {
+      "epoch": 0.3529701750061622,
+      "grad_norm": 0.18051012010020695,
+      "learning_rate": 9.670338309243738e-06,
+      "loss": 0.5205,
+      "step": 358
+    },
+    {
+      "epoch": 0.3539561252156766,
+      "grad_norm": 0.17824096043286206,
+      "learning_rate": 9.668488839454367e-06,
+      "loss": 0.5289,
+      "step": 359
+    },
+    {
+      "epoch": 0.354942075425191,
+      "grad_norm": 0.1899269535230839,
+      "learning_rate": 9.666634374078906e-06,
+      "loss": 0.5399,
+      "step": 360
+    },
+    {
+      "epoch": 0.35592802563470544,
+      "grad_norm": 0.2000896624953639,
+      "learning_rate": 9.664774915101751e-06,
+      "loss": 0.5316,
+      "step": 361
+    },
+    {
+      "epoch": 0.35691397584421986,
+      "grad_norm": 0.17982429332089273,
+      "learning_rate": 9.662910464512646e-06,
+      "loss": 0.5256,
+      "step": 362
+    },
+    {
+      "epoch": 0.3578999260537343,
+      "grad_norm": 0.18493042035953866,
+      "learning_rate": 9.661041024306673e-06,
+      "loss": 0.5602,
+      "step": 363
+    },
+    {
+      "epoch": 0.3588858762632487,
+      "grad_norm": 0.17638637049875583,
+      "learning_rate": 9.659166596484253e-06,
+      "loss": 0.5302,
+      "step": 364
+    },
+    {
+      "epoch": 0.3598718264727631,
+      "grad_norm": 0.18076729211946937,
+      "learning_rate": 9.65728718305115e-06,
+      "loss": 0.5502,
+      "step": 365
+    },
+    {
+      "epoch": 0.36085777668227753,
+      "grad_norm": 0.17444026167804935,
+      "learning_rate": 9.655402786018455e-06,
+      "loss": 0.553,
+      "step": 366
+    },
+    {
+      "epoch": 0.36184372689179195,
+      "grad_norm": 0.17009498146673663,
+      "learning_rate": 9.653513407402596e-06,
+      "loss": 0.5143,
+      "step": 367
+    },
+    {
+      "epoch": 0.36282967710130637,
+      "grad_norm": 0.21593875928225242,
+      "learning_rate": 9.651619049225328e-06,
+      "loss": 0.548,
+      "step": 368
+    },
+    {
+      "epoch": 0.3638156273108208,
+      "grad_norm": 0.1842107390428221,
+      "learning_rate": 9.649719713513742e-06,
+      "loss": 0.5366,
+      "step": 369
+    },
+    {
+      "epoch": 0.3648015775203352,
+      "grad_norm": 0.173489560161717,
+      "learning_rate": 9.647815402300247e-06,
+      "loss": 0.5181,
+      "step": 370
+    },
+    {
+      "epoch": 0.3657875277298496,
+      "grad_norm": 0.18390441935466698,
+      "learning_rate": 9.645906117622581e-06,
+      "loss": 0.5163,
+      "step": 371
+    },
+    {
+      "epoch": 0.36677347793936405,
+      "grad_norm": 0.22468546223912358,
+      "learning_rate": 9.643991861523802e-06,
+      "loss": 0.5262,
+      "step": 372
+    },
+    {
+      "epoch": 0.36775942814887846,
+      "grad_norm": 0.1900161096137712,
+      "learning_rate": 9.64207263605229e-06,
+      "loss": 0.5667,
+      "step": 373
+    },
+    {
+      "epoch": 0.3687453783583929,
+      "grad_norm": 0.18028388695730332,
+      "learning_rate": 9.640148443261739e-06,
+      "loss": 0.4991,
+      "step": 374
+    },
+    {
+      "epoch": 0.3697313285679073,
+      "grad_norm": 0.1808083614104452,
+      "learning_rate": 9.63821928521116e-06,
+      "loss": 0.5323,
+      "step": 375
+    },
+    {
+      "epoch": 0.3707172787774217,
+      "grad_norm": 0.18225409843311444,
+      "learning_rate": 9.636285163964877e-06,
+      "loss": 0.542,
+      "step": 376
+    },
+    {
+      "epoch": 0.37170322898693614,
+      "grad_norm": 0.1800024905646102,
+      "learning_rate": 9.634346081592527e-06,
+      "loss": 0.5369,
+      "step": 377
+    },
+    {
+      "epoch": 0.37268917919645056,
+      "grad_norm": 0.1878497922667102,
+      "learning_rate": 9.632402040169055e-06,
+      "loss": 0.5482,
+      "step": 378
+    },
+    {
+      "epoch": 0.373675129405965,
+      "grad_norm": 0.2620013530339256,
+      "learning_rate": 9.630453041774708e-06,
+      "loss": 0.5514,
+      "step": 379
+    },
+    {
+      "epoch": 0.3746610796154794,
+      "grad_norm": 0.18784755250804422,
+      "learning_rate": 9.628499088495043e-06,
+      "loss": 0.5375,
+      "step": 380
+    },
+    {
+      "epoch": 0.3756470298249938,
+      "grad_norm": 0.1812376528970597,
+      "learning_rate": 9.626540182420916e-06,
+      "loss": 0.5456,
+      "step": 381
+    },
+    {
+      "epoch": 0.37663298003450824,
+      "grad_norm": 0.17846141310959251,
+      "learning_rate": 9.624576325648485e-06,
+      "loss": 0.5393,
+      "step": 382
+    },
+    {
+      "epoch": 0.37761893024402265,
+      "grad_norm": 0.17766955549073946,
+      "learning_rate": 9.622607520279201e-06,
+      "loss": 0.5307,
+      "step": 383
+    },
+    {
+      "epoch": 0.3786048804535371,
+      "grad_norm": 0.1744935339508359,
+      "learning_rate": 9.620633768419819e-06,
+      "loss": 0.562,
+      "step": 384
+    },
+    {
+      "epoch": 0.3795908306630515,
+      "grad_norm": 0.18099126482002742,
+      "learning_rate": 9.618655072182376e-06,
+      "loss": 0.5515,
+      "step": 385
+    },
+    {
+      "epoch": 0.3805767808725659,
+      "grad_norm": 0.1757976737358108,
+      "learning_rate": 9.616671433684208e-06,
+      "loss": 0.5389,
+      "step": 386
+    },
+    {
+      "epoch": 0.38156273108208033,
+      "grad_norm": 0.19447865829681363,
+      "learning_rate": 9.614682855047938e-06,
+      "loss": 0.5651,
+      "step": 387
+    },
+    {
+      "epoch": 0.38254868129159475,
+      "grad_norm": 0.17219214625700813,
+      "learning_rate": 9.612689338401472e-06,
+      "loss": 0.5256,
+      "step": 388
+    },
+    {
+      "epoch": 0.38353463150110917,
+      "grad_norm": 0.17220965615808365,
+      "learning_rate": 9.610690885878002e-06,
+      "loss": 0.5323,
+      "step": 389
+    },
+    {
+      "epoch": 0.3845205817106236,
+      "grad_norm": 0.18737760870959425,
+      "learning_rate": 9.608687499616005e-06,
+      "loss": 0.5327,
+      "step": 390
+    },
+    {
+      "epoch": 0.385506531920138,
+      "grad_norm": 0.1852281145185945,
+      "learning_rate": 9.606679181759233e-06,
+      "loss": 0.5255,
+      "step": 391
+    },
+    {
+      "epoch": 0.3864924821296524,
+      "grad_norm": 0.18165566102447128,
+      "learning_rate": 9.604665934456714e-06,
+      "loss": 0.5546,
+      "step": 392
+    },
+    {
+      "epoch": 0.38747843233916685,
+      "grad_norm": 0.17861142125772,
+      "learning_rate": 9.602647759862756e-06,
+      "loss": 0.5472,
+      "step": 393
+    },
+    {
+      "epoch": 0.38846438254868126,
+      "grad_norm": 0.18792980682702926,
+      "learning_rate": 9.600624660136937e-06,
+      "loss": 0.5588,
+      "step": 394
+    },
+    {
+      "epoch": 0.38945033275819574,
+      "grad_norm": 0.18451074189767946,
+      "learning_rate": 9.598596637444101e-06,
+      "loss": 0.5338,
+      "step": 395
+    },
+    {
+      "epoch": 0.39043628296771016,
+      "grad_norm": 0.17115955881842462,
+      "learning_rate": 9.59656369395437e-06,
+      "loss": 0.5166,
+      "step": 396
+    },
+    {
+      "epoch": 0.3914222331772246,
+      "grad_norm": 0.1819762184488862,
+      "learning_rate": 9.594525831843122e-06,
+      "loss": 0.5351,
+      "step": 397
+    },
+    {
+      "epoch": 0.392408183386739,
+      "grad_norm": 0.17745975002475423,
+      "learning_rate": 9.592483053291002e-06,
+      "loss": 0.5409,
+      "step": 398
+    },
+    {
+      "epoch": 0.3933941335962534,
+      "grad_norm": 0.17283712194035614,
+      "learning_rate": 9.590435360483917e-06,
+      "loss": 0.5296,
+      "step": 399
+    },
+    {
+      "epoch": 0.39438008380576783,
+      "grad_norm": 0.1789037178138004,
+      "learning_rate": 9.588382755613029e-06,
+      "loss": 0.5506,
+      "step": 400
+    },
+    {
+      "epoch": 0.39536603401528225,
+      "grad_norm": 0.1751593495300854,
+      "learning_rate": 9.586325240874759e-06,
+      "loss": 0.5411,
+      "step": 401
+    },
+    {
+      "epoch": 0.39635198422479667,
+      "grad_norm": 0.17753137950472841,
+      "learning_rate": 9.584262818470781e-06,
+      "loss": 0.5553,
+      "step": 402
+    },
+    {
+      "epoch": 0.3973379344343111,
+      "grad_norm": 0.17729009012114222,
+      "learning_rate": 9.582195490608023e-06,
+      "loss": 0.5139,
+      "step": 403
+    },
+    {
+      "epoch": 0.3983238846438255,
+      "grad_norm": 0.17404500598409495,
+      "learning_rate": 9.580123259498658e-06,
+      "loss": 0.5367,
+      "step": 404
+    },
+    {
+      "epoch": 0.39930983485333993,
+      "grad_norm": 0.18539611447949306,
+      "learning_rate": 9.57804612736011e-06,
+      "loss": 0.5049,
+      "step": 405
+    },
+    {
+      "epoch": 0.40029578506285435,
+      "grad_norm": 0.16766267862954987,
+      "learning_rate": 9.575964096415042e-06,
+      "loss": 0.4945,
+      "step": 406
+    },
+    {
+      "epoch": 0.40128173527236877,
+      "grad_norm": 0.3041454463230651,
+      "learning_rate": 9.573877168891365e-06,
+      "loss": 0.5319,
+      "step": 407
+    },
+    {
+      "epoch": 0.4022676854818832,
+      "grad_norm": 0.1921318752247221,
+      "learning_rate": 9.571785347022225e-06,
+      "loss": 0.5181,
+      "step": 408
+    },
+    {
+      "epoch": 0.4032536356913976,
+      "grad_norm": 0.1767934643651254,
+      "learning_rate": 9.569688633046009e-06,
+      "loss": 0.5612,
+      "step": 409
+    },
+    {
+      "epoch": 0.404239585900912,
+      "grad_norm": 0.18885240184925087,
+      "learning_rate": 9.567587029206335e-06,
+      "loss": 0.5214,
+      "step": 410
+    },
+    {
+      "epoch": 0.40522553611042644,
+      "grad_norm": 0.199447723801938,
+      "learning_rate": 9.565480537752057e-06,
+      "loss": 0.5379,
+      "step": 411
+    },
+    {
+      "epoch": 0.40621148631994086,
+      "grad_norm": 0.3608200962774075,
+      "learning_rate": 9.563369160937259e-06,
+      "loss": 0.542,
+      "step": 412
+    },
+    {
+      "epoch": 0.4071974365294553,
+      "grad_norm": 0.1724817976169056,
+      "learning_rate": 9.561252901021247e-06,
+      "loss": 0.5299,
+      "step": 413
+    },
+    {
+      "epoch": 0.4081833867389697,
+      "grad_norm": 0.19258471068149405,
+      "learning_rate": 9.55913176026856e-06,
+      "loss": 0.5364,
+      "step": 414
+    },
+    {
+      "epoch": 0.4091693369484841,
+      "grad_norm": 0.1847703010304863,
+      "learning_rate": 9.557005740948954e-06,
+      "loss": 0.5597,
+      "step": 415
+    },
+    {
+      "epoch": 0.41015528715799854,
+      "grad_norm": 0.17240172918677668,
+      "learning_rate": 9.55487484533741e-06,
+      "loss": 0.5335,
+      "step": 416
+    },
+    {
+      "epoch": 0.41114123736751296,
+      "grad_norm": 0.26185902201310524,
+      "learning_rate": 9.552739075714125e-06,
+      "loss": 0.5244,
+      "step": 417
+    },
+    {
+      "epoch": 0.4121271875770274,
+      "grad_norm": 0.19211187561141485,
+      "learning_rate": 9.550598434364507e-06,
+      "loss": 0.5413,
+      "step": 418
+    },
+    {
+      "epoch": 0.4131131377865418,
+      "grad_norm": 0.1781929568483824,
+      "learning_rate": 9.548452923579186e-06,
+      "loss": 0.5464,
+      "step": 419
+    },
+    {
+      "epoch": 0.4140990879960562,
+      "grad_norm": 0.2427129508527206,
+      "learning_rate": 9.546302545653994e-06,
+      "loss": 0.544,
+      "step": 420
+    },
+    {
+      "epoch": 0.41508503820557063,
+      "grad_norm": 0.171747906030444,
+      "learning_rate": 9.544147302889977e-06,
+      "loss": 0.5144,
+      "step": 421
+    },
+    {
+      "epoch": 0.41607098841508505,
+      "grad_norm": 0.17907222130079994,
+      "learning_rate": 9.541987197593385e-06,
+      "loss": 0.536,
+      "step": 422
+    },
+    {
+      "epoch": 0.41705693862459947,
+      "grad_norm": 0.18469819930878412,
+      "learning_rate": 9.539822232075669e-06,
+      "loss": 0.5462,
+      "step": 423
+    },
+    {
+      "epoch": 0.4180428888341139,
+      "grad_norm": 0.1802968701146283,
+      "learning_rate": 9.537652408653485e-06,
+      "loss": 0.539,
+      "step": 424
+    },
+    {
+      "epoch": 0.4190288390436283,
+      "grad_norm": 0.25158217355630896,
+      "learning_rate": 9.535477729648683e-06,
+      "loss": 0.5467,
+      "step": 425
+    },
+    {
+      "epoch": 0.42001478925314273,
+      "grad_norm": 0.19205046887586966,
+      "learning_rate": 9.533298197388313e-06,
+      "loss": 0.5338,
+      "step": 426
+    },
+    {
+      "epoch": 0.42100073946265715,
+      "grad_norm": 0.17906515816988378,
+      "learning_rate": 9.531113814204611e-06,
+      "loss": 0.5444,
+      "step": 427
+    },
+    {
+      "epoch": 0.42198668967217157,
+      "grad_norm": 0.1813181727528524,
+      "learning_rate": 9.528924582435015e-06,
+      "loss": 0.5413,
+      "step": 428
+    },
+    {
+      "epoch": 0.422972639881686,
+      "grad_norm": 0.20098658554703583,
+      "learning_rate": 9.526730504422142e-06,
+      "loss": 0.542,
+      "step": 429
+    },
+    {
+      "epoch": 0.4239585900912004,
+      "grad_norm": 0.1845545628875529,
+      "learning_rate": 9.524531582513797e-06,
+      "loss": 0.5407,
+      "step": 430
+    },
+    {
+      "epoch": 0.4249445403007148,
+      "grad_norm": 0.17365774613172688,
+      "learning_rate": 9.522327819062971e-06,
+      "loss": 0.5169,
+      "step": 431
+    },
+    {
+      "epoch": 0.42593049051022924,
+      "grad_norm": 0.17486639323125394,
+      "learning_rate": 9.520119216427832e-06,
+      "loss": 0.5432,
+      "step": 432
+    },
+    {
+      "epoch": 0.42691644071974366,
+      "grad_norm": 0.19178935146833231,
+      "learning_rate": 9.517905776971731e-06,
+      "loss": 0.5269,
+      "step": 433
+    },
+    {
+      "epoch": 0.4279023909292581,
+      "grad_norm": 0.22040389128943744,
+      "learning_rate": 9.51568750306319e-06,
+      "loss": 0.5503,
+      "step": 434
+    },
+    {
+      "epoch": 0.4288883411387725,
+      "grad_norm": 0.3848732264146601,
+      "learning_rate": 9.513464397075906e-06,
+      "loss": 0.5392,
+      "step": 435
+    },
+    {
+      "epoch": 0.4298742913482869,
+      "grad_norm": 0.1800777856034973,
+      "learning_rate": 9.511236461388748e-06,
+      "loss": 0.5442,
+      "step": 436
+    },
+    {
+      "epoch": 0.43086024155780134,
+      "grad_norm": 0.17156221801999408,
+      "learning_rate": 9.509003698385751e-06,
+      "loss": 0.5259,
+      "step": 437
+    },
+    {
+      "epoch": 0.43184619176731576,
+      "grad_norm": 0.1781982902672412,
+      "learning_rate": 9.506766110456114e-06,
+      "loss": 0.5427,
+      "step": 438
+    },
+    {
+      "epoch": 0.4328321419768302,
+      "grad_norm": 0.19272411397386807,
+      "learning_rate": 9.504523699994206e-06,
+      "loss": 0.5544,
+      "step": 439
+    },
+    {
+      "epoch": 0.4338180921863446,
+      "grad_norm": 0.18322894068935697,
+      "learning_rate": 9.502276469399547e-06,
+      "loss": 0.5324,
+      "step": 440
+    },
+    {
+      "epoch": 0.434804042395859,
+      "grad_norm": 0.19942159600279194,
+      "learning_rate": 9.500024421076825e-06,
+      "loss": 0.536,
+      "step": 441
+    },
+    {
+      "epoch": 0.43578999260537343,
+      "grad_norm": 0.17674182164248262,
+      "learning_rate": 9.497767557435873e-06,
+      "loss": 0.5202,
+      "step": 442
+    },
+    {
+      "epoch": 0.43677594281488785,
+      "grad_norm": 0.1801472124058599,
+      "learning_rate": 9.495505880891683e-06,
+      "loss": 0.5397,
+      "step": 443
+    },
+    {
+      "epoch": 0.43776189302440227,
+      "grad_norm": 0.1751876020459294,
+      "learning_rate": 9.493239393864397e-06,
+      "loss": 0.5223,
+      "step": 444
+    },
+    {
+      "epoch": 0.4387478432339167,
+      "grad_norm": 0.1746613728257611,
+      "learning_rate": 9.490968098779304e-06,
+      "loss": 0.5099,
+      "step": 445
+    },
+    {
+      "epoch": 0.4397337934434311,
+      "grad_norm": 0.18029420055134318,
+      "learning_rate": 9.488691998066833e-06,
+      "loss": 0.545,
+      "step": 446
+    },
+    {
+      "epoch": 0.44071974365294553,
+      "grad_norm": 0.17599971146131657,
+      "learning_rate": 9.486411094162562e-06,
+      "loss": 0.5428,
+      "step": 447
+    },
+    {
+      "epoch": 0.44170569386245995,
+      "grad_norm": 0.17011693792438803,
+      "learning_rate": 9.484125389507206e-06,
+      "loss": 0.5267,
+      "step": 448
+    },
+    {
+      "epoch": 0.44269164407197437,
+      "grad_norm": 0.1822455591186834,
+      "learning_rate": 9.481834886546618e-06,
+      "loss": 0.5493,
+      "step": 449
+    },
+    {
+      "epoch": 0.4436775942814888,
+      "grad_norm": 0.18965504353166301,
+      "learning_rate": 9.479539587731788e-06,
+      "loss": 0.5409,
+      "step": 450
+    },
+    {
+      "epoch": 0.4446635444910032,
+      "grad_norm": 0.18495841838242305,
+      "learning_rate": 9.477239495518826e-06,
+      "loss": 0.5373,
+      "step": 451
+    },
+    {
+      "epoch": 0.4456494947005176,
+      "grad_norm": 0.17962496584208223,
+      "learning_rate": 9.474934612368989e-06,
+      "loss": 0.5243,
+      "step": 452
+    },
+    {
+      "epoch": 0.44663544491003204,
+      "grad_norm": 0.18091336056439608,
+      "learning_rate": 9.472624940748644e-06,
+      "loss": 0.5606,
+      "step": 453
+    },
+    {
+      "epoch": 0.44762139511954646,
+      "grad_norm": 0.1947142637572019,
+      "learning_rate": 9.470310483129298e-06,
+      "loss": 0.539,
+      "step": 454
+    },
+    {
+      "epoch": 0.4486073453290609,
+      "grad_norm": 0.1739883293626402,
+      "learning_rate": 9.467991241987562e-06,
+      "loss": 0.5269,
+      "step": 455
+    },
+    {
+      "epoch": 0.4495932955385753,
+      "grad_norm": 0.17548396273813394,
+      "learning_rate": 9.465667219805182e-06,
+      "loss": 0.5452,
+      "step": 456
+    },
+    {
+      "epoch": 0.4505792457480897,
+      "grad_norm": 0.17888393002480893,
+      "learning_rate": 9.463338419069007e-06,
+      "loss": 0.5415,
+      "step": 457
+    },
+    {
+      "epoch": 0.45156519595760414,
+      "grad_norm": 0.17444896541539015,
+      "learning_rate": 9.461004842271008e-06,
+      "loss": 0.5213,
+      "step": 458
+    },
+    {
+      "epoch": 0.45255114616711856,
+      "grad_norm": 0.17284634720146852,
+      "learning_rate": 9.458666491908264e-06,
+      "loss": 0.5393,
+      "step": 459
+    },
+    {
+      "epoch": 0.453537096376633,
+      "grad_norm": 0.1868956394912211,
+      "learning_rate": 9.456323370482959e-06,
+      "loss": 0.5327,
+      "step": 460
+    },
+    {
+      "epoch": 0.4545230465861474,
+      "grad_norm": 0.17853379090704663,
+      "learning_rate": 9.453975480502387e-06,
+      "loss": 0.5438,
+      "step": 461
+    },
+    {
+      "epoch": 0.4555089967956618,
+      "grad_norm": 0.1728035132474683,
+      "learning_rate": 9.451622824478941e-06,
+      "loss": 0.5396,
+      "step": 462
+    },
+    {
+      "epoch": 0.45649494700517623,
+      "grad_norm": 0.1759478823382504,
+      "learning_rate": 9.44926540493012e-06,
+      "loss": 0.5561,
+      "step": 463
+    },
+    {
+      "epoch": 0.45748089721469065,
+      "grad_norm": 0.16797933840931661,
+      "learning_rate": 9.44690322437851e-06,
+      "loss": 0.5343,
+      "step": 464
+    },
+    {
+      "epoch": 0.45846684742420507,
+      "grad_norm": 0.17310988682293296,
+      "learning_rate": 9.444536285351803e-06,
+      "loss": 0.5263,
+      "step": 465
+    },
+    {
+      "epoch": 0.4594527976337195,
+      "grad_norm": 0.1743341428430828,
+      "learning_rate": 9.442164590382771e-06,
+      "loss": 0.5234,
+      "step": 466
+    },
+    {
+      "epoch": 0.4604387478432339,
+      "grad_norm": 0.18157823661205635,
+      "learning_rate": 9.43978814200929e-06,
+      "loss": 0.5485,
+      "step": 467
+    },
+    {
+      "epoch": 0.4614246980527483,
+      "grad_norm": 0.16690352875330322,
+      "learning_rate": 9.437406942774308e-06,
+      "loss": 0.5234,
+      "step": 468
+    },
+    {
+      "epoch": 0.46241064826226275,
+      "grad_norm": 0.17294814548184448,
+      "learning_rate": 9.435020995225863e-06,
+      "loss": 0.5498,
+      "step": 469
+    },
+    {
+      "epoch": 0.46339659847177717,
+      "grad_norm": 0.18364242361789426,
+      "learning_rate": 9.432630301917075e-06,
+      "loss": 0.5321,
+      "step": 470
+    },
+    {
+      "epoch": 0.4643825486812916,
+      "grad_norm": 0.16871216430194655,
+      "learning_rate": 9.43023486540614e-06,
+      "loss": 0.5273,
+      "step": 471
+    },
+    {
+      "epoch": 0.465368498890806,
+      "grad_norm": 0.1776272651026022,
+      "learning_rate": 9.427834688256333e-06,
+      "loss": 0.5222,
+      "step": 472
+    },
+    {
+      "epoch": 0.4663544491003204,
+      "grad_norm": 0.1799030712370558,
+      "learning_rate": 9.425429773035997e-06,
+      "loss": 0.5303,
+      "step": 473
+    },
+    {
+      "epoch": 0.46734039930983484,
+      "grad_norm": 0.17383910124854074,
+      "learning_rate": 9.42302012231855e-06,
+      "loss": 0.5167,
+      "step": 474
+    },
+    {
+      "epoch": 0.46832634951934926,
+      "grad_norm": 0.19827846910160685,
+      "learning_rate": 9.420605738682471e-06,
+      "loss": 0.5152,
+      "step": 475
+    },
+    {
+      "epoch": 0.4693122997288637,
+      "grad_norm": 0.17635148231459685,
+      "learning_rate": 9.418186624711309e-06,
+      "loss": 0.54,
+      "step": 476
+    },
+    {
+      "epoch": 0.4702982499383781,
+      "grad_norm": 0.22411418332504124,
+      "learning_rate": 9.415762782993673e-06,
+      "loss": 0.5137,
+      "step": 477
+    },
+    {
+      "epoch": 0.4712842001478925,
+      "grad_norm": 0.1840318252535409,
+      "learning_rate": 9.413334216123233e-06,
+      "loss": 0.5482,
+      "step": 478
+    },
+    {
+      "epoch": 0.47227015035740694,
+      "grad_norm": 0.17458749642531488,
+      "learning_rate": 9.41090092669871e-06,
+      "loss": 0.5363,
+      "step": 479
+    },
+    {
+      "epoch": 0.47325610056692136,
+      "grad_norm": 0.17585310210282468,
+      "learning_rate": 9.408462917323882e-06,
+      "loss": 0.5117,
+      "step": 480
+    },
+    {
+      "epoch": 0.4742420507764358,
+      "grad_norm": 0.17990947019554057,
+      "learning_rate": 9.40602019060758e-06,
+      "loss": 0.5363,
+      "step": 481
+    },
+    {
+      "epoch": 0.4752280009859502,
+      "grad_norm": 0.1907254005157449,
+      "learning_rate": 9.403572749163675e-06,
+      "loss": 0.5217,
+      "step": 482
+    },
+    {
+      "epoch": 0.4762139511954646,
+      "grad_norm": 0.1821487814637426,
+      "learning_rate": 9.401120595611094e-06,
+      "loss": 0.5364,
+      "step": 483
+    },
+    {
+      "epoch": 0.47719990140497903,
+      "grad_norm": 0.18945645608652886,
+      "learning_rate": 9.398663732573798e-06,
+      "loss": 0.5305,
+      "step": 484
+    },
+    {
+      "epoch": 0.47818585161449345,
+      "grad_norm": 0.1871072551732735,
+      "learning_rate": 9.396202162680789e-06,
+      "loss": 0.532,
+      "step": 485
+    },
+    {
+      "epoch": 0.47917180182400787,
+      "grad_norm": 0.1811390463852209,
+      "learning_rate": 9.393735888566107e-06,
+      "loss": 0.5293,
+      "step": 486
+    },
+    {
+      "epoch": 0.4801577520335223,
+      "grad_norm": 0.18049327995553202,
+      "learning_rate": 9.391264912868828e-06,
+      "loss": 0.5419,
+      "step": 487
+    },
+    {
+      "epoch": 0.4811437022430367,
+      "grad_norm": 0.176874912249078,
+      "learning_rate": 9.388789238233052e-06,
+      "loss": 0.5357,
+      "step": 488
+    },
+    {
+      "epoch": 0.4821296524525511,
+      "grad_norm": 0.18204471162500285,
+      "learning_rate": 9.386308867307915e-06,
+      "loss": 0.5313,
+      "step": 489
+    },
+    {
+      "epoch": 0.48311560266206555,
+      "grad_norm": 0.20790754915654208,
+      "learning_rate": 9.383823802747572e-06,
+      "loss": 0.5152,
+      "step": 490
+    },
+    {
+      "epoch": 0.48410155287157997,
+      "grad_norm": 0.1920201424904351,
+      "learning_rate": 9.381334047211208e-06,
+      "loss": 0.5451,
+      "step": 491
+    },
+    {
+      "epoch": 0.4850875030810944,
+      "grad_norm": 0.17533154689525784,
+      "learning_rate": 9.378839603363018e-06,
+      "loss": 0.5437,
+      "step": 492
+    },
+    {
+      "epoch": 0.4860734532906088,
+      "grad_norm": 0.17772839355273848,
+      "learning_rate": 9.376340473872221e-06,
+      "loss": 0.5367,
+      "step": 493
+    },
+    {
+      "epoch": 0.4870594035001232,
+      "grad_norm": 1.094859663460843,
+      "learning_rate": 9.373836661413048e-06,
+      "loss": 0.5597,
+      "step": 494
+    },
+    {
+      "epoch": 0.48804535370963764,
+      "grad_norm": 0.18438873724873134,
+      "learning_rate": 9.37132816866474e-06,
+      "loss": 0.5432,
+      "step": 495
+    },
+    {
+      "epoch": 0.48903130391915206,
+      "grad_norm": 0.18810839204318186,
+      "learning_rate": 9.368814998311548e-06,
+      "loss": 0.5318,
+      "step": 496
+    },
+    {
+      "epoch": 0.4900172541286665,
+      "grad_norm": 0.18444665812331418,
+      "learning_rate": 9.366297153042727e-06,
+      "loss": 0.5326,
+      "step": 497
+    },
+    {
+      "epoch": 0.4910032043381809,
+      "grad_norm": 0.1707078516753226,
+      "learning_rate": 9.363774635552536e-06,
+      "loss": 0.5368,
+      "step": 498
+    },
+    {
+      "epoch": 0.4919891545476953,
+      "grad_norm": 0.18433209786482785,
+      "learning_rate": 9.36124744854023e-06,
+      "loss": 0.5153,
+      "step": 499
+    },
+    {
+      "epoch": 0.49297510475720974,
+      "grad_norm": 0.18444241059848385,
+      "learning_rate": 9.358715594710065e-06,
+      "loss": 0.5259,
+      "step": 500
+    },
+    {
+      "epoch": 0.49396105496672416,
+      "grad_norm": 0.1888150737018664,
+      "learning_rate": 9.35617907677129e-06,
+      "loss": 0.5166,
+      "step": 501
+    },
+    {
+      "epoch": 0.4949470051762386,
+      "grad_norm": 0.1865721892028227,
+      "learning_rate": 9.353637897438139e-06,
+      "loss": 0.4966,
+      "step": 502
+    },
+    {
+      "epoch": 0.495932955385753,
+      "grad_norm": 0.18872924962167356,
+      "learning_rate": 9.351092059429845e-06,
+      "loss": 0.5172,
+      "step": 503
+    },
+    {
+      "epoch": 0.4969189055952674,
+      "grad_norm": 0.17940409379970196,
+      "learning_rate": 9.348541565470614e-06,
+      "loss": 0.5352,
+      "step": 504
+    },
+    {
+      "epoch": 0.49790485580478183,
+      "grad_norm": 0.1982895133569798,
+      "learning_rate": 9.345986418289645e-06,
+      "loss": 0.526,
+      "step": 505
+    },
+    {
+      "epoch": 0.49889080601429625,
+      "grad_norm": 0.17982686539540155,
+      "learning_rate": 9.34342662062111e-06,
+      "loss": 0.5341,
+      "step": 506
+    },
+    {
+      "epoch": 0.49987675622381067,
+      "grad_norm": 0.17361686526600992,
+      "learning_rate": 9.340862175204157e-06,
+      "loss": 0.5198,
+      "step": 507
+    },
+    {
+      "epoch": 0.5008627064333251,
+      "grad_norm": 0.1753173227702939,
+      "learning_rate": 9.338293084782912e-06,
+      "loss": 0.5207,
+      "step": 508
+    },
+    {
+      "epoch": 0.5018486566428395,
+      "grad_norm": 0.1719528503287593,
+      "learning_rate": 9.335719352106465e-06,
+      "loss": 0.5261,
+      "step": 509
+    },
+    {
+      "epoch": 0.5028346068523539,
+      "grad_norm": 0.20400584078299505,
+      "learning_rate": 9.33314097992888e-06,
+      "loss": 0.5389,
+      "step": 510
+    },
+    {
+      "epoch": 0.5038205570618683,
+      "grad_norm": 0.17253150996749836,
+      "learning_rate": 9.33055797100918e-06,
+      "loss": 0.505,
+      "step": 511
+    },
+    {
+      "epoch": 0.5048065072713828,
+      "grad_norm": 0.21680273811570544,
+      "learning_rate": 9.327970328111354e-06,
+      "loss": 0.5236,
+      "step": 512
+    },
+    {
+      "epoch": 0.5057924574808972,
+      "grad_norm": 0.18983001098675922,
+      "learning_rate": 9.325378054004346e-06,
+      "loss": 0.5134,
+      "step": 513
+    },
+    {
+      "epoch": 0.5067784076904116,
+      "grad_norm": 0.23237815078364535,
+      "learning_rate": 9.32278115146206e-06,
+      "loss": 0.5373,
+      "step": 514
+    },
+    {
+      "epoch": 0.507764357899926,
+      "grad_norm": 0.18787252749594727,
+      "learning_rate": 9.32017962326335e-06,
+      "loss": 0.52,
+      "step": 515
+    },
+    {
+      "epoch": 0.5087503081094404,
+      "grad_norm": 0.1778882773289981,
+      "learning_rate": 9.317573472192018e-06,
+      "loss": 0.5191,
+      "step": 516
+    },
+    {
+      "epoch": 0.5097362583189549,
+      "grad_norm": 0.17816627152377676,
+      "learning_rate": 9.314962701036818e-06,
+      "loss": 0.5115,
+      "step": 517
+    },
+    {
+      "epoch": 0.5107222085284693,
+      "grad_norm": 0.17526363387693264,
+      "learning_rate": 9.31234731259144e-06,
+      "loss": 0.513,
+      "step": 518
+    },
+    {
+      "epoch": 0.5117081587379837,
+      "grad_norm": 0.18547672153027575,
+      "learning_rate": 9.309727309654524e-06,
+      "loss": 0.5385,
+      "step": 519
+    },
+    {
+      "epoch": 0.5126941089474981,
+      "grad_norm": 0.18888072990131097,
+      "learning_rate": 9.30710269502964e-06,
+      "loss": 0.5421,
+      "step": 520
+    },
+    {
+      "epoch": 0.5136800591570125,
+      "grad_norm": 0.1932038565992329,
+      "learning_rate": 9.3044734715253e-06,
+      "loss": 0.5524,
+      "step": 521
+    },
+    {
+      "epoch": 0.514666009366527,
+      "grad_norm": 0.17154783813009652,
+      "learning_rate": 9.301839641954937e-06,
+      "loss": 0.5404,
+      "step": 522
+    },
+    {
+      "epoch": 0.5156519595760414,
+      "grad_norm": 0.17065826544657794,
+      "learning_rate": 9.299201209136927e-06,
+      "loss": 0.5036,
+      "step": 523
+    },
+    {
+      "epoch": 0.5166379097855558,
+      "grad_norm": 0.18081266482056293,
+      "learning_rate": 9.296558175894559e-06,
+      "loss": 0.5358,
+      "step": 524
+    },
+    {
+      "epoch": 0.5176238599950702,
+      "grad_norm": 0.18409344660188967,
+      "learning_rate": 9.293910545056054e-06,
+      "loss": 0.5392,
+      "step": 525
+    },
+    {
+      "epoch": 0.5186098102045846,
+      "grad_norm": 0.18474516901228663,
+      "learning_rate": 9.291258319454546e-06,
+      "loss": 0.5325,
+      "step": 526
+    },
+    {
+      "epoch": 0.519595760414099,
+      "grad_norm": 0.16362804198311526,
+      "learning_rate": 9.28860150192809e-06,
+      "loss": 0.5267,
+      "step": 527
+    },
+    {
+      "epoch": 0.5205817106236135,
+      "grad_norm": 0.1961675315869362,
+      "learning_rate": 9.285940095319651e-06,
+      "loss": 0.5292,
+      "step": 528
+    },
+    {
+      "epoch": 0.5215676608331279,
+      "grad_norm": 0.1690370738666498,
+      "learning_rate": 9.28327410247711e-06,
+      "loss": 0.5319,
+      "step": 529
+    },
+    {
+      "epoch": 0.5225536110426423,
+      "grad_norm": 0.17344100056077083,
+      "learning_rate": 9.280603526253253e-06,
+      "loss": 0.5181,
+      "step": 530
+    },
+    {
+      "epoch": 0.5235395612521567,
+      "grad_norm": 0.1885771373904881,
+      "learning_rate": 9.277928369505766e-06,
+      "loss": 0.5287,
+      "step": 531
+    },
+    {
+      "epoch": 0.5245255114616711,
+      "grad_norm": 0.17258458207149155,
+      "learning_rate": 9.275248635097242e-06,
+      "loss": 0.5272,
+      "step": 532
+    },
+    {
+      "epoch": 0.5255114616711856,
+      "grad_norm": 0.16980858974454846,
+      "learning_rate": 9.272564325895172e-06,
+      "loss": 0.5222,
+      "step": 533
+    },
+    {
+      "epoch": 0.5264974118807,
+      "grad_norm": 0.17303373845587997,
+      "learning_rate": 9.269875444771941e-06,
+      "loss": 0.541,
+      "step": 534
+    },
+    {
+      "epoch": 0.5274833620902144,
+      "grad_norm": 0.17472353980786784,
+      "learning_rate": 9.267181994604824e-06,
+      "loss": 0.53,
+      "step": 535
+    },
+    {
+      "epoch": 0.5284693122997288,
+      "grad_norm": 0.1790405001202461,
+      "learning_rate": 9.26448397827599e-06,
+      "loss": 0.508,
+      "step": 536
+    },
+    {
+      "epoch": 0.5294552625092432,
+      "grad_norm": 0.18351127199828568,
+      "learning_rate": 9.261781398672489e-06,
+      "loss": 0.5603,
+      "step": 537
+    },
+    {
+      "epoch": 0.5304412127187577,
+      "grad_norm": 0.18363336505242986,
+      "learning_rate": 9.25907425868626e-06,
+      "loss": 0.5144,
+      "step": 538
+    },
+    {
+      "epoch": 0.5314271629282721,
+      "grad_norm": 0.17658005873610938,
+      "learning_rate": 9.256362561214116e-06,
+      "loss": 0.5267,
+      "step": 539
+    },
+    {
+      "epoch": 0.5324131131377865,
+      "grad_norm": 0.18840777055635421,
+      "learning_rate": 9.25364630915775e-06,
+      "loss": 0.5234,
+      "step": 540
+    },
+    {
+      "epoch": 0.5333990633473009,
+      "grad_norm": 0.1728518263781916,
+      "learning_rate": 9.250925505423728e-06,
+      "loss": 0.5265,
+      "step": 541
+    },
+    {
+      "epoch": 0.5343850135568153,
+      "grad_norm": 0.17488975556875405,
+      "learning_rate": 9.248200152923487e-06,
+      "loss": 0.5238,
+      "step": 542
+    },
+    {
+      "epoch": 0.5353709637663298,
+      "grad_norm": 0.1882758465967878,
+      "learning_rate": 9.24547025457333e-06,
+      "loss": 0.5564,
+      "step": 543
+    },
+    {
+      "epoch": 0.5363569139758442,
+      "grad_norm": 0.19025557740461027,
+      "learning_rate": 9.242735813294425e-06,
+      "loss": 0.5133,
+      "step": 544
+    },
+    {
+      "epoch": 0.5373428641853586,
+      "grad_norm": 0.179824214551046,
+      "learning_rate": 9.239996832012805e-06,
+      "loss": 0.5445,
+      "step": 545
+    },
+    {
+      "epoch": 0.538328814394873,
+      "grad_norm": 0.1806716065069949,
+      "learning_rate": 9.237253313659355e-06,
+      "loss": 0.5231,
+      "step": 546
+    },
+    {
+      "epoch": 0.5393147646043874,
+      "grad_norm": 0.18635574300948382,
+      "learning_rate": 9.234505261169819e-06,
+      "loss": 0.5449,
+      "step": 547
+    },
+    {
+      "epoch": 0.5403007148139019,
+      "grad_norm": 0.184656164635862,
+      "learning_rate": 9.23175267748479e-06,
+      "loss": 0.5336,
+      "step": 548
+    },
+    {
+      "epoch": 0.5412866650234163,
+      "grad_norm": 0.17411313876144827,
+      "learning_rate": 9.228995565549712e-06,
+      "loss": 0.5176,
+      "step": 549
+    },
+    {
+      "epoch": 0.5422726152329307,
+      "grad_norm": 0.1753064976401272,
+      "learning_rate": 9.226233928314874e-06,
+      "loss": 0.529,
+      "step": 550
+    },
+    {
+      "epoch": 0.5432585654424451,
+      "grad_norm": 0.20170952807937534,
+      "learning_rate": 9.223467768735407e-06,
+      "loss": 0.5297,
+      "step": 551
+    },
+    {
+      "epoch": 0.5442445156519595,
+      "grad_norm": 0.18530425670711484,
+      "learning_rate": 9.22069708977128e-06,
+      "loss": 0.54,
+      "step": 552
+    },
+    {
+      "epoch": 0.545230465861474,
+      "grad_norm": 0.20615172776597993,
+      "learning_rate": 9.217921894387303e-06,
+      "loss": 0.5094,
+      "step": 553
+    },
+    {
+      "epoch": 0.5462164160709884,
+      "grad_norm": 0.170874507407419,
+      "learning_rate": 9.21514218555311e-06,
+      "loss": 0.5132,
+      "step": 554
+    },
+    {
+      "epoch": 0.5472023662805028,
+      "grad_norm": 0.19009706896809436,
+      "learning_rate": 9.212357966243176e-06,
+      "loss": 0.5202,
+      "step": 555
+    },
+    {
+      "epoch": 0.5481883164900172,
+      "grad_norm": 0.17760598662710642,
+      "learning_rate": 9.20956923943679e-06,
+      "loss": 0.5036,
+      "step": 556
+    },
+    {
+      "epoch": 0.5491742666995316,
+      "grad_norm": 0.17949196898983627,
+      "learning_rate": 9.206776008118075e-06,
+      "loss": 0.5274,
+      "step": 557
+    },
+    {
+      "epoch": 0.550160216909046,
+      "grad_norm": 0.1693635169054839,
+      "learning_rate": 9.203978275275967e-06,
+      "loss": 0.5199,
+      "step": 558
+    },
+    {
+      "epoch": 0.5511461671185605,
+      "grad_norm": 0.17813196882293547,
+      "learning_rate": 9.201176043904225e-06,
+      "loss": 0.5278,
+      "step": 559
+    },
+    {
+      "epoch": 0.5521321173280749,
+      "grad_norm": 0.22647859583966012,
+      "learning_rate": 9.198369317001417e-06,
+      "loss": 0.5301,
+      "step": 560
+    },
+    {
+      "epoch": 0.5531180675375893,
+      "grad_norm": 0.1818989370240682,
+      "learning_rate": 9.19555809757092e-06,
+      "loss": 0.5204,
+      "step": 561
+    },
+    {
+      "epoch": 0.5541040177471037,
+      "grad_norm": 0.18242604280358907,
+      "learning_rate": 9.192742388620926e-06,
+      "loss": 0.5389,
+      "step": 562
+    },
+    {
+      "epoch": 0.5550899679566181,
+      "grad_norm": 0.17696765474750256,
+      "learning_rate": 9.189922193164424e-06,
+      "loss": 0.5295,
+      "step": 563
+    },
+    {
+      "epoch": 0.5560759181661327,
+      "grad_norm": 0.1808515728973198,
+      "learning_rate": 9.187097514219207e-06,
+      "loss": 0.5431,
+      "step": 564
+    },
+    {
+      "epoch": 0.5570618683756471,
+      "grad_norm": 0.17047869791424183,
+      "learning_rate": 9.184268354807863e-06,
+      "loss": 0.5272,
+      "step": 565
+    },
+    {
+      "epoch": 0.5580478185851615,
+      "grad_norm": 0.1900426807328964,
+      "learning_rate": 9.181434717957779e-06,
+      "loss": 0.5261,
+      "step": 566
+    },
+    {
+      "epoch": 0.5590337687946759,
+      "grad_norm": 0.17617470606974053,
+      "learning_rate": 9.17859660670113e-06,
+      "loss": 0.5311,
+      "step": 567
+    },
+    {
+      "epoch": 0.5600197190041903,
+      "grad_norm": 0.20130597373832468,
+      "learning_rate": 9.175754024074874e-06,
+      "loss": 0.5157,
+      "step": 568
+    },
+    {
+      "epoch": 0.5610056692137048,
+      "grad_norm": 0.17294261538712863,
+      "learning_rate": 9.172906973120767e-06,
+      "loss": 0.5228,
+      "step": 569
+    },
+    {
+      "epoch": 0.5619916194232192,
+      "grad_norm": 0.1852675528253617,
+      "learning_rate": 9.170055456885332e-06,
+      "loss": 0.5243,
+      "step": 570
+    },
+    {
+      "epoch": 0.5629775696327336,
+      "grad_norm": 0.18500105862976746,
+      "learning_rate": 9.16719947841988e-06,
+      "loss": 0.5399,
+      "step": 571
+    },
+    {
+      "epoch": 0.563963519842248,
+      "grad_norm": 0.1756914744676518,
+      "learning_rate": 9.164339040780492e-06,
+      "loss": 0.5188,
+      "step": 572
+    },
+    {
+      "epoch": 0.5649494700517624,
+      "grad_norm": 0.2025007793819143,
+      "learning_rate": 9.161474147028022e-06,
+      "loss": 0.5283,
+      "step": 573
+    },
+    {
+      "epoch": 0.5659354202612769,
+      "grad_norm": 0.1808833374255632,
+      "learning_rate": 9.158604800228094e-06,
+      "loss": 0.5186,
+      "step": 574
+    },
+    {
+      "epoch": 0.5669213704707913,
+      "grad_norm": 0.1867472385948832,
+      "learning_rate": 9.155731003451091e-06,
+      "loss": 0.5013,
+      "step": 575
+    },
+    {
+      "epoch": 0.5679073206803057,
+      "grad_norm": 0.18062805229765697,
+      "learning_rate": 9.152852759772167e-06,
+      "loss": 0.5165,
+      "step": 576
+    },
+    {
+      "epoch": 0.5688932708898201,
+      "grad_norm": 0.17552623211314305,
+      "learning_rate": 9.149970072271226e-06,
+      "loss": 0.5034,
+      "step": 577
+    },
+    {
+      "epoch": 0.5698792210993345,
+      "grad_norm": 0.1919350685661946,
+      "learning_rate": 9.147082944032934e-06,
+      "loss": 0.5172,
+      "step": 578
+    },
+    {
+      "epoch": 0.570865171308849,
+      "grad_norm": 0.1811007654992155,
+      "learning_rate": 9.144191378146702e-06,
+      "loss": 0.5199,
+      "step": 579
+    },
+    {
+      "epoch": 0.5718511215183634,
+      "grad_norm": 0.19407623548069264,
+      "learning_rate": 9.141295377706695e-06,
+      "loss": 0.527,
+      "step": 580
+    },
+    {
+      "epoch": 0.5728370717278778,
+      "grad_norm": 0.18908186768753915,
+      "learning_rate": 9.138394945811825e-06,
+      "loss": 0.5129,
+      "step": 581
+    },
+    {
+      "epoch": 0.5738230219373922,
+      "grad_norm": 0.18119416709811703,
+      "learning_rate": 9.135490085565735e-06,
+      "loss": 0.5419,
+      "step": 582
+    },
+    {
+      "epoch": 0.5748089721469066,
+      "grad_norm": 0.17536626522865337,
+      "learning_rate": 9.132580800076819e-06,
+      "loss": 0.5419,
+      "step": 583
+    },
+    {
+      "epoch": 0.575794922356421,
+      "grad_norm": 0.17746659463522532,
+      "learning_rate": 9.129667092458198e-06,
+      "loss": 0.5174,
+      "step": 584
+    },
+    {
+      "epoch": 0.5767808725659355,
+      "grad_norm": 0.1857542289742682,
+      "learning_rate": 9.126748965827732e-06,
+      "loss": 0.5035,
+      "step": 585
+    },
+    {
+      "epoch": 0.5777668227754499,
+      "grad_norm": 0.17427240686444098,
+      "learning_rate": 9.123826423308005e-06,
+      "loss": 0.5292,
+      "step": 586
+    },
+    {
+      "epoch": 0.5787527729849643,
+      "grad_norm": 0.18841192624205355,
+      "learning_rate": 9.120899468026327e-06,
+      "loss": 0.5384,
+      "step": 587
+    },
+    {
+      "epoch": 0.5797387231944787,
+      "grad_norm": 0.18603041739827225,
+      "learning_rate": 9.117968103114729e-06,
+      "loss": 0.5434,
+      "step": 588
+    },
+    {
+      "epoch": 0.5807246734039931,
+      "grad_norm": 0.17510196686929236,
+      "learning_rate": 9.115032331709961e-06,
+      "loss": 0.5213,
+      "step": 589
+    },
+    {
+      "epoch": 0.5817106236135076,
+      "grad_norm": 0.1747927492804185,
+      "learning_rate": 9.112092156953493e-06,
+      "loss": 0.529,
+      "step": 590
+    },
+    {
+      "epoch": 0.582696573823022,
+      "grad_norm": 0.17858183539184375,
+      "learning_rate": 9.109147581991499e-06,
+      "loss": 0.5252,
+      "step": 591
+    },
+    {
+      "epoch": 0.5836825240325364,
+      "grad_norm": 0.17461562316476606,
+      "learning_rate": 9.106198609974865e-06,
+      "loss": 0.5383,
+      "step": 592
+    },
+    {
+      "epoch": 0.5846684742420508,
+      "grad_norm": 0.18330748062867017,
+      "learning_rate": 9.103245244059187e-06,
+      "loss": 0.5146,
+      "step": 593
+    },
+    {
+      "epoch": 0.5856544244515652,
+      "grad_norm": 0.1753610355571392,
+      "learning_rate": 9.100287487404753e-06,
+      "loss": 0.5159,
+      "step": 594
+    },
+    {
+      "epoch": 0.5866403746610797,
+      "grad_norm": 0.17367460310203844,
+      "learning_rate": 9.097325343176558e-06,
+      "loss": 0.5065,
+      "step": 595
+    },
+    {
+      "epoch": 0.5876263248705941,
+      "grad_norm": 0.17222303467067504,
+      "learning_rate": 9.094358814544286e-06,
+      "loss": 0.5376,
+      "step": 596
+    },
+    {
+      "epoch": 0.5886122750801085,
+      "grad_norm": 0.1873340504183569,
+      "learning_rate": 9.091387904682318e-06,
+      "loss": 0.5202,
+      "step": 597
+    },
+    {
+      "epoch": 0.5895982252896229,
+      "grad_norm": 0.17438861375145898,
+      "learning_rate": 9.08841261676972e-06,
+      "loss": 0.5212,
+      "step": 598
+    },
+    {
+      "epoch": 0.5905841754991373,
+      "grad_norm": 0.1714457753169691,
+      "learning_rate": 9.08543295399024e-06,
+      "loss": 0.5055,
+      "step": 599
+    },
+    {
+      "epoch": 0.5915701257086518,
+      "grad_norm": 0.20022070999928204,
+      "learning_rate": 9.082448919532314e-06,
+      "loss": 0.4946,
+      "step": 600
+    },
+    {
+      "epoch": 0.5925560759181662,
+      "grad_norm": 0.178991294093694,
+      "learning_rate": 9.07946051658905e-06,
+      "loss": 0.5387,
+      "step": 601
+    },
+    {
+      "epoch": 0.5935420261276806,
+      "grad_norm": 0.18384867168921457,
+      "learning_rate": 9.076467748358237e-06,
+      "loss": 0.542,
+      "step": 602
+    },
+    {
+      "epoch": 0.594527976337195,
+      "grad_norm": 0.1725702905420044,
+      "learning_rate": 9.073470618042328e-06,
+      "loss": 0.5228,
+      "step": 603
+    },
+    {
+      "epoch": 0.5955139265467094,
+      "grad_norm": 0.16585800029266956,
+      "learning_rate": 9.070469128848447e-06,
+      "loss": 0.5035,
+      "step": 604
+    },
+    {
+      "epoch": 0.5964998767562238,
+      "grad_norm": 0.5047088349732528,
+      "learning_rate": 9.067463283988385e-06,
+      "loss": 0.5234,
+      "step": 605
+    },
+    {
+      "epoch": 0.5974858269657383,
+      "grad_norm": 0.17584146068307602,
+      "learning_rate": 9.064453086678587e-06,
+      "loss": 0.5461,
+      "step": 606
+    },
+    {
+      "epoch": 0.5984717771752527,
+      "grad_norm": 0.20307560060614097,
+      "learning_rate": 9.061438540140161e-06,
+      "loss": 0.5296,
+      "step": 607
+    },
+    {
+      "epoch": 0.5994577273847671,
+      "grad_norm": 0.17851206137289966,
+      "learning_rate": 9.05841964759887e-06,
+      "loss": 0.5365,
+      "step": 608
+    },
+    {
+      "epoch": 0.6004436775942815,
+      "grad_norm": 0.17546432011636895,
+      "learning_rate": 9.055396412285122e-06,
+      "loss": 0.5211,
+      "step": 609
+    },
+    {
+      "epoch": 0.6014296278037959,
+      "grad_norm": 0.17844165754682811,
+      "learning_rate": 9.052368837433978e-06,
+      "loss": 0.52,
+      "step": 610
+    },
+    {
+      "epoch": 0.6024155780133104,
+      "grad_norm": 0.1687356005922332,
+      "learning_rate": 9.049336926285136e-06,
+      "loss": 0.516,
+      "step": 611
+    },
+    {
+      "epoch": 0.6034015282228248,
+      "grad_norm": 0.18488627019817228,
+      "learning_rate": 9.04630068208294e-06,
+      "loss": 0.5448,
+      "step": 612
+    },
+    {
+      "epoch": 0.6043874784323392,
+      "grad_norm": 0.17424086572152608,
+      "learning_rate": 9.043260108076369e-06,
+      "loss": 0.5281,
+      "step": 613
+    },
+    {
+      "epoch": 0.6053734286418536,
+      "grad_norm": 0.17573018876393326,
+      "learning_rate": 9.040215207519031e-06,
+      "loss": 0.5125,
+      "step": 614
+    },
+    {
+      "epoch": 0.606359378851368,
+      "grad_norm": 0.17297871673089396,
+      "learning_rate": 9.037165983669172e-06,
+      "loss": 0.5335,
+      "step": 615
+    },
+    {
+      "epoch": 0.6073453290608825,
+      "grad_norm": 0.18495495026508482,
+      "learning_rate": 9.034112439789655e-06,
+      "loss": 0.5165,
+      "step": 616
+    },
+    {
+      "epoch": 0.6083312792703969,
+      "grad_norm": 0.18458607108773536,
+      "learning_rate": 9.031054579147973e-06,
+      "loss": 0.5222,
+      "step": 617
+    },
+    {
+      "epoch": 0.6093172294799113,
+      "grad_norm": 0.18087886436501827,
+      "learning_rate": 9.027992405016234e-06,
+      "loss": 0.5323,
+      "step": 618
+    },
+    {
+      "epoch": 0.6103031796894257,
+      "grad_norm": 0.1801917170488805,
+      "learning_rate": 9.024925920671167e-06,
+      "loss": 0.5426,
+      "step": 619
+    },
+    {
+      "epoch": 0.6112891298989401,
+      "grad_norm": 0.17289962272813006,
+      "learning_rate": 9.021855129394103e-06,
+      "loss": 0.528,
+      "step": 620
+    },
+    {
+      "epoch": 0.6122750801084546,
+      "grad_norm": 0.17531792307433441,
+      "learning_rate": 9.018780034470991e-06,
+      "loss": 0.5216,
+      "step": 621
+    },
+    {
+      "epoch": 0.613261030317969,
+      "grad_norm": 0.1741054805237634,
+      "learning_rate": 9.015700639192384e-06,
+      "loss": 0.5243,
+      "step": 622
+    },
+    {
+      "epoch": 0.6142469805274834,
+      "grad_norm": 0.1835618139824688,
+      "learning_rate": 9.012616946853432e-06,
+      "loss": 0.5287,
+      "step": 623
+    },
+    {
+      "epoch": 0.6152329307369978,
+      "grad_norm": 0.17638053998512027,
+      "learning_rate": 9.009528960753885e-06,
+      "loss": 0.5331,
+      "step": 624
+    },
+    {
+      "epoch": 0.6162188809465122,
+      "grad_norm": 0.17975674286395826,
+      "learning_rate": 9.006436684198093e-06,
+      "loss": 0.4977,
+      "step": 625
+    },
+    {
+      "epoch": 0.6172048311560266,
+      "grad_norm": 0.18533891500734204,
+      "learning_rate": 9.00334012049499e-06,
+      "loss": 0.5129,
+      "step": 626
+    },
+    {
+      "epoch": 0.6181907813655411,
+      "grad_norm": 0.18383024671858167,
+      "learning_rate": 9.0002392729581e-06,
+      "loss": 0.5275,
+      "step": 627
+    },
+    {
+      "epoch": 0.6191767315750555,
+      "grad_norm": 0.16821043921604145,
+      "learning_rate": 8.99713414490553e-06,
+      "loss": 0.4997,
+      "step": 628
+    },
+    {
+      "epoch": 0.6201626817845699,
+      "grad_norm": 0.1704480999997968,
+      "learning_rate": 8.994024739659972e-06,
+      "loss": 0.5139,
+      "step": 629
+    },
+    {
+      "epoch": 0.6211486319940843,
+      "grad_norm": 0.18746952704693728,
+      "learning_rate": 8.990911060548689e-06,
+      "loss": 0.5194,
+      "step": 630
+    },
+    {
+      "epoch": 0.6221345822035987,
+      "grad_norm": 0.17341484319326284,
+      "learning_rate": 8.98779311090352e-06,
+      "loss": 0.5233,
+      "step": 631
+    },
+    {
+      "epoch": 0.6231205324131132,
+      "grad_norm": 0.17844220202436867,
+      "learning_rate": 8.984670894060874e-06,
+      "loss": 0.5087,
+      "step": 632
+    },
+    {
+      "epoch": 0.6241064826226276,
+      "grad_norm": 0.16855815973950497,
+      "learning_rate": 8.981544413361725e-06,
+      "loss": 0.5104,
+      "step": 633
+    },
+    {
+      "epoch": 0.625092432832142,
+      "grad_norm": 0.1849169795066938,
+      "learning_rate": 8.978413672151612e-06,
+      "loss": 0.5251,
+      "step": 634
+    },
+    {
+      "epoch": 0.6260783830416564,
+      "grad_norm": 0.1773493478914938,
+      "learning_rate": 8.97527867378063e-06,
+      "loss": 0.5235,
+      "step": 635
+    },
+    {
+      "epoch": 0.6270643332511708,
+      "grad_norm": 0.1841968528665142,
+      "learning_rate": 8.97213942160343e-06,
+      "loss": 0.529,
+      "step": 636
+    },
+    {
+      "epoch": 0.6280502834606853,
+      "grad_norm": 0.17046732498513104,
+      "learning_rate": 8.968995918979216e-06,
+      "loss": 0.5131,
+      "step": 637
+    },
+    {
+      "epoch": 0.6290362336701997,
+      "grad_norm": 0.17779413754694615,
+      "learning_rate": 8.96584816927174e-06,
+      "loss": 0.5082,
+      "step": 638
+    },
+    {
+      "epoch": 0.6300221838797141,
+      "grad_norm": 0.1770139560344267,
+      "learning_rate": 8.962696175849299e-06,
+      "loss": 0.5194,
+      "step": 639
+    },
+    {
+      "epoch": 0.6310081340892285,
+      "grad_norm": 0.23355583624196918,
+      "learning_rate": 8.959539942084731e-06,
+      "loss": 0.5152,
+      "step": 640
+    },
+    {
+      "epoch": 0.6319940842987429,
+      "grad_norm": 0.18163116429500578,
+      "learning_rate": 8.956379471355411e-06,
+      "loss": 0.5312,
+      "step": 641
+    },
+    {
+      "epoch": 0.6329800345082573,
+      "grad_norm": 0.17497411266633903,
+      "learning_rate": 8.953214767043246e-06,
+      "loss": 0.5166,
+      "step": 642
+    },
+    {
+      "epoch": 0.6339659847177718,
+      "grad_norm": 0.17616177689851764,
+      "learning_rate": 8.950045832534678e-06,
+      "loss": 0.5048,
+      "step": 643
+    },
+    {
+      "epoch": 0.6349519349272862,
+      "grad_norm": 0.17784544913925193,
+      "learning_rate": 8.946872671220669e-06,
+      "loss": 0.5136,
+      "step": 644
+    },
+    {
+      "epoch": 0.6359378851368006,
+      "grad_norm": 0.17088153319930033,
+      "learning_rate": 8.943695286496709e-06,
+      "loss": 0.4895,
+      "step": 645
+    },
+    {
+      "epoch": 0.636923835346315,
+      "grad_norm": 0.17832756103730774,
+      "learning_rate": 8.940513681762807e-06,
+      "loss": 0.5211,
+      "step": 646
+    },
+    {
+      "epoch": 0.6379097855558294,
+      "grad_norm": 0.17537595009721352,
+      "learning_rate": 8.937327860423487e-06,
+      "loss": 0.5225,
+      "step": 647
+    },
+    {
+      "epoch": 0.6388957357653439,
+      "grad_norm": 0.18717683211393102,
+      "learning_rate": 8.93413782588778e-06,
+      "loss": 0.5315,
+      "step": 648
+    },
+    {
+      "epoch": 0.6398816859748583,
+      "grad_norm": 0.25977649967137717,
+      "learning_rate": 8.930943581569236e-06,
+      "loss": 0.5431,
+      "step": 649
+    },
+    {
+      "epoch": 0.6408676361843727,
+      "grad_norm": 0.1800160217835687,
+      "learning_rate": 8.927745130885902e-06,
+      "loss": 0.5173,
+      "step": 650
+    },
+    {
+      "epoch": 0.6418535863938871,
+      "grad_norm": 0.1896989176209276,
+      "learning_rate": 8.924542477260325e-06,
+      "loss": 0.5168,
+      "step": 651
+    },
+    {
+      "epoch": 0.6428395366034015,
+      "grad_norm": 0.17510675412492296,
+      "learning_rate": 8.921335624119555e-06,
+      "loss": 0.5334,
+      "step": 652
+    },
+    {
+      "epoch": 0.643825486812916,
+      "grad_norm": 0.1846458906517486,
+      "learning_rate": 8.918124574895133e-06,
+      "loss": 0.5108,
+      "step": 653
+    },
+    {
+      "epoch": 0.6448114370224304,
+      "grad_norm": 0.18874230582806142,
+      "learning_rate": 8.91490933302309e-06,
+      "loss": 0.5174,
+      "step": 654
+    },
+    {
+      "epoch": 0.6457973872319448,
+      "grad_norm": 0.19499657504014306,
+      "learning_rate": 8.911689901943943e-06,
+      "loss": 0.5125,
+      "step": 655
+    },
+    {
+      "epoch": 0.6467833374414592,
+      "grad_norm": 0.18253219610497026,
+      "learning_rate": 8.90846628510269e-06,
+      "loss": 0.508,
+      "step": 656
+    },
+    {
+      "epoch": 0.6477692876509736,
+      "grad_norm": 0.1744591790741951,
+      "learning_rate": 8.905238485948815e-06,
+      "loss": 0.5352,
+      "step": 657
+    },
+    {
+      "epoch": 0.648755237860488,
+      "grad_norm": 0.18783949825896712,
+      "learning_rate": 8.90200650793627e-06,
+      "loss": 0.5306,
+      "step": 658
+    },
+    {
+      "epoch": 0.6497411880700025,
+      "grad_norm": 0.17134031494103244,
+      "learning_rate": 8.89877035452348e-06,
+      "loss": 0.5014,
+      "step": 659
+    },
+    {
+      "epoch": 0.6507271382795169,
+      "grad_norm": 0.17370655934781326,
+      "learning_rate": 8.895530029173343e-06,
+      "loss": 0.5102,
+      "step": 660
+    },
+    {
+      "epoch": 0.6517130884890313,
+      "grad_norm": 0.1770600492093783,
+      "learning_rate": 8.892285535353212e-06,
+      "loss": 0.5314,
+      "step": 661
+    },
+    {
+      "epoch": 0.6526990386985457,
+      "grad_norm": 0.18204208757325732,
+      "learning_rate": 8.889036876534911e-06,
+      "loss": 0.5348,
+      "step": 662
+    },
+    {
+      "epoch": 0.6536849889080601,
+      "grad_norm": 0.17762241359082767,
+      "learning_rate": 8.885784056194712e-06,
+      "loss": 0.5296,
+      "step": 663
+    },
+    {
+      "epoch": 0.6546709391175746,
+      "grad_norm": 0.1676215990546039,
+      "learning_rate": 8.882527077813348e-06,
+      "loss": 0.4964,
+      "step": 664
+    },
+    {
+      "epoch": 0.655656889327089,
+      "grad_norm": 0.1882492905393916,
+      "learning_rate": 8.879265944875994e-06,
+      "loss": 0.5196,
+      "step": 665
+    },
+    {
+      "epoch": 0.6566428395366034,
+      "grad_norm": 0.17397694941479602,
+      "learning_rate": 8.876000660872274e-06,
+      "loss": 0.5111,
+      "step": 666
+    },
+    {
+      "epoch": 0.6576287897461178,
+      "grad_norm": 0.17999310124671788,
+      "learning_rate": 8.872731229296256e-06,
+      "loss": 0.5402,
+      "step": 667
+    },
+    {
+      "epoch": 0.6586147399556322,
+      "grad_norm": 0.16951200375869988,
+      "learning_rate": 8.869457653646443e-06,
+      "loss": 0.5375,
+      "step": 668
+    },
+    {
+      "epoch": 0.6596006901651467,
+      "grad_norm": 0.18029882359082766,
+      "learning_rate": 8.866179937425772e-06,
+      "loss": 0.5394,
+      "step": 669
+    },
+    {
+      "epoch": 0.6605866403746611,
+      "grad_norm": 0.1869965519670085,
+      "learning_rate": 8.862898084141614e-06,
+      "loss": 0.5185,
+      "step": 670
+    },
+    {
+      "epoch": 0.6615725905841755,
+      "grad_norm": 0.16817396755280328,
+      "learning_rate": 8.859612097305764e-06,
+      "loss": 0.5186,
+      "step": 671
+    },
+    {
+      "epoch": 0.6625585407936899,
+      "grad_norm": 0.17438563993837336,
+      "learning_rate": 8.85632198043444e-06,
+      "loss": 0.5272,
+      "step": 672
+    },
+    {
+      "epoch": 0.6635444910032043,
+      "grad_norm": 0.1804169563087905,
+      "learning_rate": 8.853027737048286e-06,
+      "loss": 0.507,
+      "step": 673
+    },
+    {
+      "epoch": 0.6645304412127188,
+      "grad_norm": 0.1785577926904514,
+      "learning_rate": 8.849729370672352e-06,
+      "loss": 0.5215,
+      "step": 674
+    },
+    {
+      "epoch": 0.6655163914222332,
+      "grad_norm": 0.16911658733520982,
+      "learning_rate": 8.8464268848361e-06,
+      "loss": 0.5232,
+      "step": 675
+    },
+    {
+      "epoch": 0.6665023416317476,
+      "grad_norm": 0.1773874030449928,
+      "learning_rate": 8.843120283073415e-06,
+      "loss": 0.531,
+      "step": 676
+    },
+    {
+      "epoch": 0.667488291841262,
+      "grad_norm": 0.397460651171539,
+      "learning_rate": 8.839809568922565e-06,
+      "loss": 0.5075,
+      "step": 677
+    },
+    {
+      "epoch": 0.6684742420507764,
+      "grad_norm": 0.18509188624181927,
+      "learning_rate": 8.836494745926234e-06,
+      "loss": 0.5245,
+      "step": 678
+    },
+    {
+      "epoch": 0.6694601922602909,
+      "grad_norm": 0.17677712795862216,
+      "learning_rate": 8.833175817631499e-06,
+      "loss": 0.5132,
+      "step": 679
+    },
+    {
+      "epoch": 0.6704461424698053,
+      "grad_norm": 0.1915713550121279,
+      "learning_rate": 8.829852787589824e-06,
+      "loss": 0.5186,
+      "step": 680
+    },
+    {
+      "epoch": 0.6714320926793197,
+      "grad_norm": 0.22366661643432598,
+      "learning_rate": 8.826525659357071e-06,
+      "loss": 0.5316,
+      "step": 681
+    },
+    {
+      "epoch": 0.6724180428888341,
+      "grad_norm": 0.17281416028542135,
+      "learning_rate": 8.823194436493483e-06,
+      "loss": 0.4999,
+      "step": 682
+    },
+    {
+      "epoch": 0.6734039930983485,
+      "grad_norm": 0.2658779409346162,
+      "learning_rate": 8.819859122563682e-06,
+      "loss": 0.5118,
+      "step": 683
+    },
+    {
+      "epoch": 0.674389943307863,
+      "grad_norm": 0.18389819905434257,
+      "learning_rate": 8.816519721136673e-06,
+      "loss": 0.5159,
+      "step": 684
+    },
+    {
+      "epoch": 0.6753758935173774,
+      "grad_norm": 0.1845332508473855,
+      "learning_rate": 8.813176235785829e-06,
+      "loss": 0.5409,
+      "step": 685
+    },
+    {
+      "epoch": 0.6763618437268918,
+      "grad_norm": 0.17160868622207487,
+      "learning_rate": 8.8098286700889e-06,
+      "loss": 0.5265,
+      "step": 686
+    },
+    {
+      "epoch": 0.6773477939364062,
+      "grad_norm": 0.17548251551800362,
+      "learning_rate": 8.806477027627997e-06,
+      "loss": 0.5121,
+      "step": 687
+    },
+    {
+      "epoch": 0.6783337441459206,
+      "grad_norm": 0.17118384896055586,
+      "learning_rate": 8.803121311989598e-06,
+      "loss": 0.5327,
+      "step": 688
+    },
+    {
+      "epoch": 0.679319694355435,
+      "grad_norm": 0.171583103142981,
+      "learning_rate": 8.79976152676453e-06,
+      "loss": 0.5305,
+      "step": 689
+    },
+    {
+      "epoch": 0.6803056445649495,
+      "grad_norm": 0.17345039764589554,
+      "learning_rate": 8.796397675547986e-06,
+      "loss": 0.5304,
+      "step": 690
+    },
+    {
+      "epoch": 0.6812915947744639,
+      "grad_norm": 0.16964715352228815,
+      "learning_rate": 8.793029761939504e-06,
+      "loss": 0.5196,
+      "step": 691
+    },
+    {
+      "epoch": 0.6822775449839783,
+      "grad_norm": 0.19762274913482514,
+      "learning_rate": 8.789657789542972e-06,
+      "loss": 0.5103,
+      "step": 692
+    },
+    {
+      "epoch": 0.6832634951934927,
+      "grad_norm": 0.20395516806477976,
+      "learning_rate": 8.786281761966615e-06,
+      "loss": 0.5261,
+      "step": 693
+    },
+    {
+      "epoch": 0.6842494454030071,
+      "grad_norm": 0.18121458134398816,
+      "learning_rate": 8.782901682823004e-06,
+      "loss": 0.5254,
+      "step": 694
+    },
+    {
+      "epoch": 0.6852353956125216,
+      "grad_norm": 0.16959004669484334,
+      "learning_rate": 8.779517555729046e-06,
+      "loss": 0.5196,
+      "step": 695
+    },
+    {
+      "epoch": 0.686221345822036,
+      "grad_norm": 0.1845303297736926,
+      "learning_rate": 8.776129384305973e-06,
+      "loss": 0.5282,
+      "step": 696
+    },
+    {
+      "epoch": 0.6872072960315504,
+      "grad_norm": 0.1759390137413818,
+      "learning_rate": 8.772737172179348e-06,
+      "loss": 0.5129,
+      "step": 697
+    },
+    {
+      "epoch": 0.6881932462410648,
+      "grad_norm": 0.1689420049611807,
+      "learning_rate": 8.769340922979062e-06,
+      "loss": 0.5357,
+      "step": 698
+    },
+    {
+      "epoch": 0.6891791964505792,
+      "grad_norm": 0.18054094111426847,
+      "learning_rate": 8.765940640339318e-06,
+      "loss": 0.5182,
+      "step": 699
+    },
+    {
+      "epoch": 0.6901651466600937,
+      "grad_norm": 0.1678558299690107,
+      "learning_rate": 8.76253632789864e-06,
+      "loss": 0.5313,
+      "step": 700
+    },
+    {
+      "epoch": 0.6911510968696081,
+      "grad_norm": 0.17772178459431145,
+      "learning_rate": 8.759127989299865e-06,
+      "loss": 0.5093,
+      "step": 701
+    },
+    {
+      "epoch": 0.6921370470791225,
+      "grad_norm": 0.21720545871461938,
+      "learning_rate": 8.755715628190136e-06,
+      "loss": 0.5381,
+      "step": 702
+    },
+    {
+      "epoch": 0.6931229972886369,
+      "grad_norm": 0.17115182309046265,
+      "learning_rate": 8.752299248220901e-06,
+      "loss": 0.504,
+      "step": 703
+    },
+    {
+      "epoch": 0.6941089474981513,
+      "grad_norm": 0.18270723154242413,
+      "learning_rate": 8.748878853047906e-06,
+      "loss": 0.519,
+      "step": 704
+    },
+    {
+      "epoch": 0.6950948977076657,
+      "grad_norm": 0.1702302493339554,
+      "learning_rate": 8.7454544463312e-06,
+      "loss": 0.5246,
+      "step": 705
+    },
+    {
+      "epoch": 0.6960808479171802,
+      "grad_norm": 0.19264316148857302,
+      "learning_rate": 8.742026031735116e-06,
+      "loss": 0.5322,
+      "step": 706
+    },
+    {
+      "epoch": 0.6970667981266946,
+      "grad_norm": 0.17481732405788628,
+      "learning_rate": 8.738593612928283e-06,
+      "loss": 0.5046,
+      "step": 707
+    },
+    {
+      "epoch": 0.698052748336209,
+      "grad_norm": 0.17351857644214455,
+      "learning_rate": 8.735157193583611e-06,
+      "loss": 0.5307,
+      "step": 708
+    },
+    {
+      "epoch": 0.6990386985457234,
+      "grad_norm": 0.32734345905410767,
+      "learning_rate": 8.731716777378292e-06,
+      "loss": 0.5222,
+      "step": 709
+    },
+    {
+      "epoch": 0.7000246487552378,
+      "grad_norm": 0.18202642303027347,
+      "learning_rate": 8.728272367993795e-06,
+      "loss": 0.5131,
+      "step": 710
+    },
+    {
+      "epoch": 0.7010105989647523,
+      "grad_norm": 0.17044576847981274,
+      "learning_rate": 8.724823969115861e-06,
+      "loss": 0.5321,
+      "step": 711
+    },
+    {
+      "epoch": 0.7019965491742667,
+      "grad_norm": 0.18106248863996377,
+      "learning_rate": 8.721371584434502e-06,
+      "loss": 0.5256,
+      "step": 712
+    },
+    {
+      "epoch": 0.7029824993837811,
+      "grad_norm": 0.1824247086532533,
+      "learning_rate": 8.71791521764399e-06,
+      "loss": 0.5178,
+      "step": 713
+    },
+    {
+      "epoch": 0.7039684495932955,
+      "grad_norm": 0.16428313462842645,
+      "learning_rate": 8.714454872442869e-06,
+      "loss": 0.5167,
+      "step": 714
+    },
+    {
+      "epoch": 0.7049543998028099,
+      "grad_norm": 0.17087528650372252,
+      "learning_rate": 8.710990552533923e-06,
+      "loss": 0.5316,
+      "step": 715
+    },
+    {
+      "epoch": 0.7059403500123244,
+      "grad_norm": 0.17731853379074156,
+      "learning_rate": 8.707522261624208e-06,
+      "loss": 0.5168,
+      "step": 716
+    },
+    {
+      "epoch": 0.7069263002218388,
+      "grad_norm": 0.18217115811783596,
+      "learning_rate": 8.704050003425015e-06,
+      "loss": 0.5193,
+      "step": 717
+    },
+    {
+      "epoch": 0.7079122504313532,
+      "grad_norm": 0.17094344995648159,
+      "learning_rate": 8.700573781651889e-06,
+      "loss": 0.5213,
+      "step": 718
+    },
+    {
+      "epoch": 0.7088982006408676,
+      "grad_norm": 0.17574602574233736,
+      "learning_rate": 8.69709360002461e-06,
+      "loss": 0.5281,
+      "step": 719
+    },
+    {
+      "epoch": 0.709884150850382,
+      "grad_norm": 0.17241572672508582,
+      "learning_rate": 8.693609462267202e-06,
+      "loss": 0.5147,
+      "step": 720
+    },
+    {
+      "epoch": 0.7108701010598965,
+      "grad_norm": 0.18222696431042285,
+      "learning_rate": 8.690121372107914e-06,
+      "loss": 0.5169,
+      "step": 721
+    },
+    {
+      "epoch": 0.7118560512694109,
+      "grad_norm": 0.17050122835486803,
+      "learning_rate": 8.686629333279233e-06,
+      "loss": 0.5161,
+      "step": 722
+    },
+    {
+      "epoch": 0.7128420014789253,
+      "grad_norm": 0.5388747937922178,
+      "learning_rate": 8.683133349517863e-06,
+      "loss": 0.52,
+      "step": 723
+    },
+    {
+      "epoch": 0.7138279516884397,
+      "grad_norm": 0.1680344800860356,
+      "learning_rate": 8.679633424564739e-06,
+      "loss": 0.5104,
+      "step": 724
+    },
+    {
+      "epoch": 0.7148139018979541,
+      "grad_norm": 0.17414658048357418,
+      "learning_rate": 8.676129562165004e-06,
+      "loss": 0.5249,
+      "step": 725
+    },
+    {
+      "epoch": 0.7157998521074685,
+      "grad_norm": 0.17932147298985338,
+      "learning_rate": 8.672621766068017e-06,
+      "loss": 0.5428,
+      "step": 726
+    },
+    {
+      "epoch": 0.716785802316983,
+      "grad_norm": 0.16685313372848734,
+      "learning_rate": 8.66911004002735e-06,
+      "loss": 0.523,
+      "step": 727
+    },
+    {
+      "epoch": 0.7177717525264974,
+      "grad_norm": 0.1877324601151875,
+      "learning_rate": 8.66559438780078e-06,
+      "loss": 0.5054,
+      "step": 728
+    },
+    {
+      "epoch": 0.7187577027360118,
+      "grad_norm": 0.16681356492773824,
+      "learning_rate": 8.66207481315028e-06,
+      "loss": 0.5048,
+      "step": 729
+    },
+    {
+      "epoch": 0.7197436529455262,
+      "grad_norm": 0.1805457431900985,
+      "learning_rate": 8.658551319842022e-06,
+      "loss": 0.529,
+      "step": 730
+    },
+    {
+      "epoch": 0.7207296031550406,
+      "grad_norm": 0.17751202813071432,
+      "learning_rate": 8.655023911646374e-06,
+      "loss": 0.5418,
+      "step": 731
+    },
+    {
+      "epoch": 0.7217155533645551,
+      "grad_norm": 0.17157275389362875,
+      "learning_rate": 8.651492592337895e-06,
+      "loss": 0.5143,
+      "step": 732
+    },
+    {
+      "epoch": 0.7227015035740695,
+      "grad_norm": 0.18539140436717147,
+      "learning_rate": 8.647957365695321e-06,
+      "loss": 0.5189,
+      "step": 733
+    },
+    {
+      "epoch": 0.7236874537835839,
+      "grad_norm": 0.16796283507271556,
+      "learning_rate": 8.644418235501576e-06,
+      "loss": 0.5215,
+      "step": 734
+    },
+    {
+      "epoch": 0.7246734039930983,
+      "grad_norm": 0.1747485196836248,
+      "learning_rate": 8.64087520554376e-06,
+      "loss": 0.5347,
+      "step": 735
+    },
+    {
+      "epoch": 0.7256593542026127,
+      "grad_norm": 0.3195405598301076,
+      "learning_rate": 8.637328279613143e-06,
+      "loss": 0.5253,
+      "step": 736
+    },
+    {
+      "epoch": 0.7266453044121272,
+      "grad_norm": 0.17565048687872226,
+      "learning_rate": 8.633777461505167e-06,
+      "loss": 0.5154,
+      "step": 737
+    },
+    {
+      "epoch": 0.7276312546216416,
+      "grad_norm": 0.1733570592289281,
+      "learning_rate": 8.630222755019437e-06,
+      "loss": 0.502,
+      "step": 738
+    },
+    {
+      "epoch": 0.728617204831156,
+      "grad_norm": 0.3577630848232857,
+      "learning_rate": 8.626664163959722e-06,
+      "loss": 0.5404,
+      "step": 739
+    },
+    {
+      "epoch": 0.7296031550406704,
+      "grad_norm": 0.17605315644053063,
+      "learning_rate": 8.623101692133943e-06,
+      "loss": 0.4973,
+      "step": 740
+    },
+    {
+      "epoch": 0.7305891052501848,
+      "grad_norm": 0.17239188746088818,
+      "learning_rate": 8.61953534335418e-06,
+      "loss": 0.5219,
+      "step": 741
+    },
+    {
+      "epoch": 0.7315750554596993,
+      "grad_norm": 0.17111425984714337,
+      "learning_rate": 8.615965121436652e-06,
+      "loss": 0.5205,
+      "step": 742
+    },
+    {
+      "epoch": 0.7325610056692137,
+      "grad_norm": 0.20812294181276428,
+      "learning_rate": 8.61239103020173e-06,
+      "loss": 0.5341,
+      "step": 743
+    },
+    {
+      "epoch": 0.7335469558787281,
+      "grad_norm": 0.17875989631698133,
+      "learning_rate": 8.608813073473927e-06,
+      "loss": 0.5249,
+      "step": 744
+    },
+    {
+      "epoch": 0.7345329060882425,
+      "grad_norm": 0.17716741956150453,
+      "learning_rate": 8.605231255081885e-06,
+      "loss": 0.5377,
+      "step": 745
+    },
+    {
+      "epoch": 0.7355188562977569,
+      "grad_norm": 0.16697367846409525,
+      "learning_rate": 8.601645578858385e-06,
+      "loss": 0.5362,
+      "step": 746
+    },
+    {
+      "epoch": 0.7365048065072713,
+      "grad_norm": 0.16740450652200117,
+      "learning_rate": 8.598056048640331e-06,
+      "loss": 0.516,
+      "step": 747
+    },
+    {
+      "epoch": 0.7374907567167858,
+      "grad_norm": 0.17036646832008162,
+      "learning_rate": 8.594462668268754e-06,
+      "loss": 0.5254,
+      "step": 748
+    },
+    {
+      "epoch": 0.7384767069263002,
+      "grad_norm": 0.18306208877356162,
+      "learning_rate": 8.590865441588804e-06,
+      "loss": 0.5133,
+      "step": 749
+    },
+    {
+      "epoch": 0.7394626571358146,
+      "grad_norm": 0.16885245950372,
+      "learning_rate": 8.58726437244975e-06,
+      "loss": 0.5193,
+      "step": 750
+    },
+    {
+      "epoch": 0.740448607345329,
+      "grad_norm": 0.1762670178152663,
+      "learning_rate": 8.583659464704965e-06,
+      "loss": 0.5338,
+      "step": 751
+    },
+    {
+      "epoch": 0.7414345575548434,
+      "grad_norm": 0.17222086219642968,
+      "learning_rate": 8.580050722211937e-06,
+      "loss": 0.5196,
+      "step": 752
+    },
+    {
+      "epoch": 0.7424205077643579,
+      "grad_norm": 0.17374197299439567,
+      "learning_rate": 8.576438148832256e-06,
+      "loss": 0.5273,
+      "step": 753
+    },
+    {
+      "epoch": 0.7434064579738723,
+      "grad_norm": 0.17175106736459506,
+      "learning_rate": 8.572821748431606e-06,
+      "loss": 0.5182,
+      "step": 754
+    },
+    {
+      "epoch": 0.7443924081833867,
+      "grad_norm": 0.1706566026586807,
+      "learning_rate": 8.569201524879775e-06,
+      "loss": 0.5236,
+      "step": 755
+    },
+    {
+      "epoch": 0.7453783583929011,
+      "grad_norm": 0.17427390185507804,
+      "learning_rate": 8.565577482050631e-06,
+      "loss": 0.5223,
+      "step": 756
+    },
+    {
+      "epoch": 0.7463643086024155,
+      "grad_norm": 0.18022321211747883,
+      "learning_rate": 8.561949623822141e-06,
+      "loss": 0.5225,
+      "step": 757
+    },
+    {
+      "epoch": 0.74735025881193,
+      "grad_norm": 0.17662547715213833,
+      "learning_rate": 8.558317954076349e-06,
+      "loss": 0.5166,
+      "step": 758
+    },
+    {
+      "epoch": 0.7483362090214444,
+      "grad_norm": 0.17273233750102043,
+      "learning_rate": 8.554682476699372e-06,
+      "loss": 0.5381,
+      "step": 759
+    },
+    {
+      "epoch": 0.7493221592309588,
+      "grad_norm": 0.1872194206953377,
+      "learning_rate": 8.55104319558141e-06,
+      "loss": 0.5323,
+      "step": 760
+    },
+    {
+      "epoch": 0.7503081094404732,
+      "grad_norm": 0.2527148861263636,
+      "learning_rate": 8.547400114616728e-06,
+      "loss": 0.5125,
+      "step": 761
+    },
+    {
+      "epoch": 0.7512940596499876,
+      "grad_norm": 0.17465398129673773,
+      "learning_rate": 8.543753237703665e-06,
+      "loss": 0.5248,
+      "step": 762
+    },
+    {
+      "epoch": 0.752280009859502,
+      "grad_norm": 0.1660106297923328,
+      "learning_rate": 8.540102568744608e-06,
+      "loss": 0.5109,
+      "step": 763
+    },
+    {
+      "epoch": 0.7532659600690165,
+      "grad_norm": 0.17253706651070183,
+      "learning_rate": 8.536448111646017e-06,
+      "loss": 0.5307,
+      "step": 764
+    },
+    {
+      "epoch": 0.7542519102785309,
+      "grad_norm": 0.17289757333017136,
+      "learning_rate": 8.532789870318392e-06,
+      "loss": 0.5077,
+      "step": 765
+    },
+    {
+      "epoch": 0.7552378604880453,
+      "grad_norm": 0.1868838629606392,
+      "learning_rate": 8.529127848676293e-06,
+      "loss": 0.5228,
+      "step": 766
+    },
+    {
+      "epoch": 0.7562238106975597,
+      "grad_norm": 0.18131235865759035,
+      "learning_rate": 8.525462050638317e-06,
+      "loss": 0.5178,
+      "step": 767
+    },
+    {
+      "epoch": 0.7572097609070741,
+      "grad_norm": 0.17232552266199797,
+      "learning_rate": 8.521792480127111e-06,
+      "loss": 0.5171,
+      "step": 768
+    },
+    {
+      "epoch": 0.7581957111165886,
+      "grad_norm": 0.1746415429343793,
+      "learning_rate": 8.51811914106935e-06,
+      "loss": 0.5182,
+      "step": 769
+    },
+    {
+      "epoch": 0.759181661326103,
+      "grad_norm": 0.17857413236160963,
+      "learning_rate": 8.514442037395747e-06,
+      "loss": 0.5148,
+      "step": 770
+    },
+    {
+      "epoch": 0.7601676115356174,
+      "grad_norm": 0.17572257432702684,
+      "learning_rate": 8.51076117304104e-06,
+      "loss": 0.5189,
+      "step": 771
+    },
+    {
+      "epoch": 0.7611535617451318,
+      "grad_norm": 0.16957297368102137,
+      "learning_rate": 8.507076551943993e-06,
+      "loss": 0.5157,
+      "step": 772
+    },
+    {
+      "epoch": 0.7621395119546462,
+      "grad_norm": 0.18128339179208697,
+      "learning_rate": 8.503388178047392e-06,
+      "loss": 0.53,
+      "step": 773
+    },
+    {
+      "epoch": 0.7631254621641607,
+      "grad_norm": 0.16976166249276048,
+      "learning_rate": 8.499696055298033e-06,
+      "loss": 0.5065,
+      "step": 774
+    },
+    {
+      "epoch": 0.7641114123736751,
+      "grad_norm": 0.2109153690670825,
+      "learning_rate": 8.496000187646729e-06,
+      "loss": 0.5076,
+      "step": 775
+    },
+    {
+      "epoch": 0.7650973625831895,
+      "grad_norm": 0.17292177920516993,
+      "learning_rate": 8.4923005790483e-06,
+      "loss": 0.504,
+      "step": 776
+    },
+    {
+      "epoch": 0.7660833127927039,
+      "grad_norm": 0.20856268456176974,
+      "learning_rate": 8.488597233461563e-06,
+      "loss": 0.5301,
+      "step": 777
+    },
+    {
+      "epoch": 0.7670692630022183,
+      "grad_norm": 0.16833157301414384,
+      "learning_rate": 8.48489015484934e-06,
+      "loss": 0.5145,
+      "step": 778
+    },
+    {
+      "epoch": 0.7680552132117328,
+      "grad_norm": 0.17347222595579526,
+      "learning_rate": 8.48117934717845e-06,
+      "loss": 0.516,
+      "step": 779
+    },
+    {
+      "epoch": 0.7690411634212472,
+      "grad_norm": 0.24132985412414357,
+      "learning_rate": 8.47746481441969e-06,
+      "loss": 0.5063,
+      "step": 780
+    },
+    {
+      "epoch": 0.7700271136307616,
+      "grad_norm": 0.17026387255922254,
+      "learning_rate": 8.473746560547857e-06,
+      "loss": 0.5172,
+      "step": 781
+    },
+    {
+      "epoch": 0.771013063840276,
+      "grad_norm": 0.16638691748823983,
+      "learning_rate": 8.470024589541724e-06,
+      "loss": 0.5201,
+      "step": 782
+    },
+    {
+      "epoch": 0.7719990140497904,
+      "grad_norm": 0.36498027012790685,
+      "learning_rate": 8.466298905384039e-06,
+      "loss": 0.5063,
+      "step": 783
+    },
+    {
+      "epoch": 0.7729849642593049,
+      "grad_norm": 0.16813785042319146,
+      "learning_rate": 8.462569512061526e-06,
+      "loss": 0.5365,
+      "step": 784
+    },
+    {
+      "epoch": 0.7739709144688193,
+      "grad_norm": 0.17121635835533072,
+      "learning_rate": 8.458836413564881e-06,
+      "loss": 0.5179,
+      "step": 785
+    },
+    {
+      "epoch": 0.7749568646783337,
+      "grad_norm": 0.17475909353692448,
+      "learning_rate": 8.45509961388876e-06,
+      "loss": 0.5146,
+      "step": 786
+    },
+    {
+      "epoch": 0.7759428148878481,
+      "grad_norm": 0.2162109626239601,
+      "learning_rate": 8.451359117031779e-06,
+      "loss": 0.528,
+      "step": 787
+    },
+    {
+      "epoch": 0.7769287650973625,
+      "grad_norm": 0.18007170603621503,
+      "learning_rate": 8.447614926996513e-06,
+      "loss": 0.5175,
+      "step": 788
+    },
+    {
+      "epoch": 0.7779147153068771,
+      "grad_norm": 0.17144065048679238,
+      "learning_rate": 8.443867047789494e-06,
+      "loss": 0.5148,
+      "step": 789
+    },
+    {
+      "epoch": 0.7789006655163915,
+      "grad_norm": 0.17851513494300253,
+      "learning_rate": 8.440115483421187e-06,
+      "loss": 0.5311,
+      "step": 790
+    },
+    {
+      "epoch": 0.7798866157259059,
+      "grad_norm": 0.16612173862337312,
+      "learning_rate": 8.436360237906017e-06,
+      "loss": 0.512,
+      "step": 791
+    },
+    {
+      "epoch": 0.7808725659354203,
+      "grad_norm": 0.1853966280987676,
+      "learning_rate": 8.432601315262336e-06,
+      "loss": 0.5019,
+      "step": 792
+    },
+    {
+      "epoch": 0.7818585161449347,
+      "grad_norm": 0.17683658205732966,
+      "learning_rate": 8.428838719512437e-06,
+      "loss": 0.484,
+      "step": 793
+    },
+    {
+      "epoch": 0.7828444663544492,
+      "grad_norm": 0.536136400317554,
+      "learning_rate": 8.425072454682543e-06,
+      "loss": 0.5322,
+      "step": 794
+    },
+    {
+      "epoch": 0.7838304165639636,
+      "grad_norm": 0.28768019540866874,
+      "learning_rate": 8.421302524802799e-06,
+      "loss": 0.5279,
+      "step": 795
+    },
+    {
+      "epoch": 0.784816366773478,
+      "grad_norm": 0.16627609196370421,
+      "learning_rate": 8.417528933907276e-06,
+      "loss": 0.5076,
+      "step": 796
+    },
+    {
+      "epoch": 0.7858023169829924,
+      "grad_norm": 0.17247981442273191,
+      "learning_rate": 8.413751686033961e-06,
+      "loss": 0.5199,
+      "step": 797
+    },
+    {
+      "epoch": 0.7867882671925068,
+      "grad_norm": 0.17103119472700135,
+      "learning_rate": 8.409970785224755e-06,
+      "loss": 0.5059,
+      "step": 798
+    },
+    {
+      "epoch": 0.7877742174020212,
+      "grad_norm": 0.18970462671162505,
+      "learning_rate": 8.406186235525466e-06,
+      "loss": 0.514,
+      "step": 799
+    },
+    {
+      "epoch": 0.7887601676115357,
+      "grad_norm": 0.17506332006283146,
+      "learning_rate": 8.402398040985809e-06,
+      "loss": 0.5157,
+      "step": 800
+    },
+    {
+      "epoch": 0.7897461178210501,
+      "grad_norm": 0.17254204827600841,
+      "learning_rate": 8.398606205659397e-06,
+      "loss": 0.5093,
+      "step": 801
+    },
+    {
+      "epoch": 0.7907320680305645,
+      "grad_norm": 0.1715289364194175,
+      "learning_rate": 8.394810733603742e-06,
+      "loss": 0.5174,
+      "step": 802
+    },
+    {
+      "epoch": 0.7917180182400789,
+      "grad_norm": 0.16717707688301195,
+      "learning_rate": 8.391011628880243e-06,
+      "loss": 0.5077,
+      "step": 803
+    },
+    {
+      "epoch": 0.7927039684495933,
+      "grad_norm": 0.18976171328191369,
+      "learning_rate": 8.387208895554191e-06,
+      "loss": 0.5261,
+      "step": 804
+    },
+    {
+      "epoch": 0.7936899186591078,
+      "grad_norm": 0.18091511950181954,
+      "learning_rate": 8.383402537694755e-06,
+      "loss": 0.5329,
+      "step": 805
+    },
+    {
+      "epoch": 0.7946758688686222,
+      "grad_norm": 0.18000960716089773,
+      "learning_rate": 8.379592559374987e-06,
+      "loss": 0.5244,
+      "step": 806
+    },
+    {
+      "epoch": 0.7956618190781366,
+      "grad_norm": 0.1713765859795502,
+      "learning_rate": 8.37577896467181e-06,
+      "loss": 0.5069,
+      "step": 807
+    },
+    {
+      "epoch": 0.796647769287651,
+      "grad_norm": 0.1676906230612834,
+      "learning_rate": 8.371961757666018e-06,
+      "loss": 0.5138,
+      "step": 808
+    },
+    {
+      "epoch": 0.7976337194971654,
+      "grad_norm": 0.17398939308947162,
+      "learning_rate": 8.36814094244227e-06,
+      "loss": 0.5008,
+      "step": 809
+    },
+    {
+      "epoch": 0.7986196697066799,
+      "grad_norm": 0.16647096267780487,
+      "learning_rate": 8.364316523089089e-06,
+      "loss": 0.5075,
+      "step": 810
+    },
+    {
+      "epoch": 0.7996056199161943,
+      "grad_norm": 0.17850069609849975,
+      "learning_rate": 8.360488503698848e-06,
+      "loss": 0.5286,
+      "step": 811
+    },
+    {
+      "epoch": 0.8005915701257087,
+      "grad_norm": 0.17037329244238567,
+      "learning_rate": 8.35665688836778e-06,
+      "loss": 0.5076,
+      "step": 812
+    },
+    {
+      "epoch": 0.8015775203352231,
+      "grad_norm": 0.17934894999043896,
+      "learning_rate": 8.352821681195958e-06,
+      "loss": 0.4978,
+      "step": 813
+    },
+    {
+      "epoch": 0.8025634705447375,
+      "grad_norm": 0.17546299537076732,
+      "learning_rate": 8.348982886287305e-06,
+      "loss": 0.5278,
+      "step": 814
+    },
+    {
+      "epoch": 0.803549420754252,
+      "grad_norm": 0.1703902987238829,
+      "learning_rate": 8.345140507749579e-06,
+      "loss": 0.5226,
+      "step": 815
+    },
+    {
+      "epoch": 0.8045353709637664,
+      "grad_norm": 0.1684317982131377,
+      "learning_rate": 8.341294549694379e-06,
+      "loss": 0.5147,
+      "step": 816
+    },
+    {
+      "epoch": 0.8055213211732808,
+      "grad_norm": 0.17252078349532643,
+      "learning_rate": 8.337445016237124e-06,
+      "loss": 0.5209,
+      "step": 817
+    },
+    {
+      "epoch": 0.8065072713827952,
+      "grad_norm": 0.17367668306172382,
+      "learning_rate": 8.333591911497069e-06,
+      "loss": 0.5316,
+      "step": 818
+    },
+    {
+      "epoch": 0.8074932215923096,
+      "grad_norm": 0.16782970866036928,
+      "learning_rate": 8.329735239597282e-06,
+      "loss": 0.504,
+      "step": 819
+    },
+    {
+      "epoch": 0.808479171801824,
+      "grad_norm": 0.17483998670023923,
+      "learning_rate": 8.325875004664659e-06,
+      "loss": 0.5135,
+      "step": 820
+    },
+    {
+      "epoch": 0.8094651220113385,
+      "grad_norm": 0.1708654556051924,
+      "learning_rate": 8.322011210829895e-06,
+      "loss": 0.5333,
+      "step": 821
+    },
+    {
+      "epoch": 0.8104510722208529,
+      "grad_norm": 0.17013008093470075,
+      "learning_rate": 8.318143862227504e-06,
+      "loss": 0.5073,
+      "step": 822
+    },
+    {
+      "epoch": 0.8114370224303673,
+      "grad_norm": 0.16511648195326395,
+      "learning_rate": 8.314272962995796e-06,
+      "loss": 0.5136,
+      "step": 823
+    },
+    {
+      "epoch": 0.8124229726398817,
+      "grad_norm": 0.16893211551503326,
+      "learning_rate": 8.31039851727689e-06,
+      "loss": 0.4973,
+      "step": 824
+    },
+    {
+      "epoch": 0.8134089228493961,
+      "grad_norm": 0.17199240095995227,
+      "learning_rate": 8.30652052921669e-06,
+      "loss": 0.5096,
+      "step": 825
+    },
+    {
+      "epoch": 0.8143948730589106,
+      "grad_norm": 0.18280485626898793,
+      "learning_rate": 8.302639002964899e-06,
+      "loss": 0.5188,
+      "step": 826
+    },
+    {
+      "epoch": 0.815380823268425,
+      "grad_norm": 0.16706532361316342,
+      "learning_rate": 8.298753942674999e-06,
+      "loss": 0.5116,
+      "step": 827
+    },
+    {
+      "epoch": 0.8163667734779394,
+      "grad_norm": 0.1721919933141353,
+      "learning_rate": 8.294865352504257e-06,
+      "loss": 0.5346,
+      "step": 828
+    },
+    {
+      "epoch": 0.8173527236874538,
+      "grad_norm": 0.17172107926109706,
+      "learning_rate": 8.290973236613718e-06,
+      "loss": 0.5342,
+      "step": 829
+    },
+    {
+      "epoch": 0.8183386738969682,
+      "grad_norm": 0.21013618327969716,
+      "learning_rate": 8.287077599168197e-06,
+      "loss": 0.524,
+      "step": 830
+    },
+    {
+      "epoch": 0.8193246241064827,
+      "grad_norm": 0.17027795744853666,
+      "learning_rate": 8.283178444336281e-06,
+      "loss": 0.5326,
+      "step": 831
+    },
+    {
+      "epoch": 0.8203105743159971,
+      "grad_norm": 0.17343975258847022,
+      "learning_rate": 8.279275776290316e-06,
+      "loss": 0.5108,
+      "step": 832
+    },
+    {
+      "epoch": 0.8212965245255115,
+      "grad_norm": 0.1712135908466209,
+      "learning_rate": 8.275369599206415e-06,
+      "loss": 0.5173,
+      "step": 833
+    },
+    {
+      "epoch": 0.8222824747350259,
+      "grad_norm": 0.18313799778430767,
+      "learning_rate": 8.271459917264435e-06,
+      "loss": 0.5245,
+      "step": 834
+    },
+    {
+      "epoch": 0.8232684249445403,
+      "grad_norm": 0.17493574661254696,
+      "learning_rate": 8.267546734647993e-06,
+      "loss": 0.5171,
+      "step": 835
+    },
+    {
+      "epoch": 0.8242543751540548,
+      "grad_norm": 0.16832759951679077,
+      "learning_rate": 8.263630055544447e-06,
+      "loss": 0.5179,
+      "step": 836
+    },
+    {
+      "epoch": 0.8252403253635692,
+      "grad_norm": 0.16866467872444635,
+      "learning_rate": 8.2597098841449e-06,
+      "loss": 0.5068,
+      "step": 837
+    },
+    {
+      "epoch": 0.8262262755730836,
+      "grad_norm": 0.1808763535003984,
+      "learning_rate": 8.25578622464419e-06,
+      "loss": 0.4934,
+      "step": 838
+    },
+    {
+      "epoch": 0.827212225782598,
+      "grad_norm": 0.1655924359407733,
+      "learning_rate": 8.251859081240882e-06,
+      "loss": 0.5008,
+      "step": 839
+    },
+    {
+      "epoch": 0.8281981759921124,
+      "grad_norm": 0.1794886277416319,
+      "learning_rate": 8.24792845813728e-06,
+      "loss": 0.5236,
+      "step": 840
+    },
+    {
+      "epoch": 0.8291841262016268,
+      "grad_norm": 0.18082788296448596,
+      "learning_rate": 8.243994359539404e-06,
+      "loss": 0.5277,
+      "step": 841
+    },
+    {
+      "epoch": 0.8301700764111413,
+      "grad_norm": 0.17032445693735299,
+      "learning_rate": 8.240056789656996e-06,
+      "loss": 0.492,
+      "step": 842
+    },
+    {
+      "epoch": 0.8311560266206557,
+      "grad_norm": 0.1818549810915091,
+      "learning_rate": 8.23611575270351e-06,
+      "loss": 0.5179,
+      "step": 843
+    },
+    {
+      "epoch": 0.8321419768301701,
+      "grad_norm": 0.1800975573391804,
+      "learning_rate": 8.23217125289611e-06,
+      "loss": 0.5441,
+      "step": 844
+    },
+    {
+      "epoch": 0.8331279270396845,
+      "grad_norm": 0.17835570335483725,
+      "learning_rate": 8.228223294455668e-06,
+      "loss": 0.5355,
+      "step": 845
+    },
+    {
+      "epoch": 0.8341138772491989,
+      "grad_norm": 0.1786970391841637,
+      "learning_rate": 8.224271881606758e-06,
+      "loss": 0.5266,
+      "step": 846
+    },
+    {
+      "epoch": 0.8350998274587134,
+      "grad_norm": 0.1719155152862329,
+      "learning_rate": 8.220317018577645e-06,
+      "loss": 0.5048,
+      "step": 847
+    },
+    {
+      "epoch": 0.8360857776682278,
+      "grad_norm": 0.1919660610537843,
+      "learning_rate": 8.216358709600291e-06,
+      "loss": 0.5093,
+      "step": 848
+    },
+    {
+      "epoch": 0.8370717278777422,
+      "grad_norm": 0.2929155996744783,
+      "learning_rate": 8.212396958910343e-06,
+      "loss": 0.5188,
+      "step": 849
+    },
+    {
+      "epoch": 0.8380576780872566,
+      "grad_norm": 0.16517385158404066,
+      "learning_rate": 8.208431770747133e-06,
+      "loss": 0.5163,
+      "step": 850
+    },
+    {
+      "epoch": 0.839043628296771,
+      "grad_norm": 0.1610551985539912,
+      "learning_rate": 8.204463149353667e-06,
+      "loss": 0.5115,
+      "step": 851
+    },
+    {
+      "epoch": 0.8400295785062855,
+      "grad_norm": 0.17113931041207608,
+      "learning_rate": 8.20049109897663e-06,
+      "loss": 0.5045,
+      "step": 852
+    },
+    {
+      "epoch": 0.8410155287157999,
+      "grad_norm": 0.17732907409269819,
+      "learning_rate": 8.19651562386637e-06,
+      "loss": 0.5134,
+      "step": 853
+    },
+    {
+      "epoch": 0.8420014789253143,
+      "grad_norm": 0.16174127147822506,
+      "learning_rate": 8.192536728276907e-06,
+      "loss": 0.5275,
+      "step": 854
+    },
+    {
+      "epoch": 0.8429874291348287,
+      "grad_norm": 0.18285945322027478,
+      "learning_rate": 8.188554416465918e-06,
+      "loss": 0.5036,
+      "step": 855
+    },
+    {
+      "epoch": 0.8439733793443431,
+      "grad_norm": 0.17393897848244694,
+      "learning_rate": 8.184568692694732e-06,
+      "loss": 0.5335,
+      "step": 856
+    },
+    {
+      "epoch": 0.8449593295538576,
+      "grad_norm": 0.16875773096341296,
+      "learning_rate": 8.180579561228334e-06,
+      "loss": 0.5056,
+      "step": 857
+    },
+    {
+      "epoch": 0.845945279763372,
+      "grad_norm": 0.1708121787085384,
+      "learning_rate": 8.176587026335354e-06,
+      "loss": 0.5163,
+      "step": 858
+    },
+    {
+      "epoch": 0.8469312299728864,
+      "grad_norm": 0.18144067908380584,
+      "learning_rate": 8.172591092288062e-06,
+      "loss": 0.5174,
+      "step": 859
+    },
+    {
+      "epoch": 0.8479171801824008,
+      "grad_norm": 0.17396213489252815,
+      "learning_rate": 8.168591763362369e-06,
+      "loss": 0.532,
+      "step": 860
+    },
+    {
+      "epoch": 0.8489031303919152,
+      "grad_norm": 0.17377129128530383,
+      "learning_rate": 8.164589043837814e-06,
+      "loss": 0.51,
+      "step": 861
+    },
+    {
+      "epoch": 0.8498890806014296,
+      "grad_norm": 0.17497985775285094,
+      "learning_rate": 8.160582937997567e-06,
+      "loss": 0.5287,
+      "step": 862
+    },
+    {
+      "epoch": 0.8508750308109441,
+      "grad_norm": 0.17391851185726995,
+      "learning_rate": 8.156573450128425e-06,
+      "loss": 0.5177,
+      "step": 863
+    },
+    {
+      "epoch": 0.8518609810204585,
+      "grad_norm": 0.3233822011352339,
+      "learning_rate": 8.152560584520794e-06,
+      "loss": 0.5058,
+      "step": 864
+    },
+    {
+      "epoch": 0.8528469312299729,
+      "grad_norm": 0.18345526720652805,
+      "learning_rate": 8.148544345468707e-06,
+      "loss": 0.5169,
+      "step": 865
+    },
+    {
+      "epoch": 0.8538328814394873,
+      "grad_norm": 0.18520161454156156,
+      "learning_rate": 8.144524737269797e-06,
+      "loss": 0.5204,
+      "step": 866
+    },
+    {
+      "epoch": 0.8548188316490017,
+      "grad_norm": 0.16136672291650989,
+      "learning_rate": 8.140501764225304e-06,
+      "loss": 0.4984,
+      "step": 867
+    },
+    {
+      "epoch": 0.8558047818585162,
+      "grad_norm": 0.17069620807049551,
+      "learning_rate": 8.136475430640076e-06,
+      "loss": 0.5251,
+      "step": 868
+    },
+    {
+      "epoch": 0.8567907320680306,
+      "grad_norm": 0.17689553735324884,
+      "learning_rate": 8.132445740822546e-06,
+      "loss": 0.5151,
+      "step": 869
+    },
+    {
+      "epoch": 0.857776682277545,
+      "grad_norm": 0.18881379637189347,
+      "learning_rate": 8.128412699084744e-06,
+      "loss": 0.5173,
+      "step": 870
+    },
+    {
+      "epoch": 0.8587626324870594,
+      "grad_norm": 0.17266036667964668,
+      "learning_rate": 8.12437630974229e-06,
+      "loss": 0.504,
+      "step": 871
+    },
+    {
+      "epoch": 0.8597485826965738,
+      "grad_norm": 0.17363778574025837,
+      "learning_rate": 8.120336577114382e-06,
+      "loss": 0.5288,
+      "step": 872
+    },
+    {
+      "epoch": 0.8607345329060883,
+      "grad_norm": 0.17300665468675847,
+      "learning_rate": 8.116293505523793e-06,
+      "loss": 0.5192,
+      "step": 873
+    },
+    {
+      "epoch": 0.8617204831156027,
+      "grad_norm": 0.16167858955319575,
+      "learning_rate": 8.112247099296873e-06,
+      "loss": 0.5215,
+      "step": 874
+    },
+    {
+      "epoch": 0.8627064333251171,
+      "grad_norm": 0.16165411639123678,
+      "learning_rate": 8.108197362763542e-06,
+      "loss": 0.4974,
+      "step": 875
+    },
+    {
+      "epoch": 0.8636923835346315,
+      "grad_norm": 0.17052900436983048,
+      "learning_rate": 8.104144300257277e-06,
+      "loss": 0.5027,
+      "step": 876
+    },
+    {
+      "epoch": 0.8646783337441459,
+      "grad_norm": 0.170918407922491,
+      "learning_rate": 8.100087916115121e-06,
+      "loss": 0.5205,
+      "step": 877
+    },
+    {
+      "epoch": 0.8656642839536604,
+      "grad_norm": 0.1658525751356633,
+      "learning_rate": 8.096028214677666e-06,
+      "loss": 0.4945,
+      "step": 878
+    },
+    {
+      "epoch": 0.8666502341631748,
+      "grad_norm": 0.1638429649649223,
+      "learning_rate": 8.09196520028906e-06,
+      "loss": 0.4963,
+      "step": 879
+    },
+    {
+      "epoch": 0.8676361843726892,
+      "grad_norm": 0.17112577903597162,
+      "learning_rate": 8.08789887729699e-06,
+      "loss": 0.5219,
+      "step": 880
+    },
+    {
+      "epoch": 0.8686221345822036,
+      "grad_norm": 0.17209000636907013,
+      "learning_rate": 8.08382925005268e-06,
+      "loss": 0.5191,
+      "step": 881
+    },
+    {
+      "epoch": 0.869608084791718,
+      "grad_norm": 0.17224830148926035,
+      "learning_rate": 8.079756322910906e-06,
+      "loss": 0.5275,
+      "step": 882
+    },
+    {
+      "epoch": 0.8705940350012324,
+      "grad_norm": 0.1736943438173366,
+      "learning_rate": 8.075680100229957e-06,
+      "loss": 0.5165,
+      "step": 883
+    },
+    {
+      "epoch": 0.8715799852107469,
+      "grad_norm": 0.18029154396746958,
+      "learning_rate": 8.071600586371655e-06,
+      "loss": 0.5139,
+      "step": 884
+    },
+    {
+      "epoch": 0.8725659354202613,
+      "grad_norm": 0.17136245587960056,
+      "learning_rate": 8.067517785701352e-06,
+      "loss": 0.5233,
+      "step": 885
+    },
+    {
+      "epoch": 0.8735518856297757,
+      "grad_norm": 0.17555967885359414,
+      "learning_rate": 8.0634317025879e-06,
+      "loss": 0.5443,
+      "step": 886
+    },
+    {
+      "epoch": 0.8745378358392901,
+      "grad_norm": 0.16247925338321514,
+      "learning_rate": 8.059342341403683e-06,
+      "loss": 0.5183,
+      "step": 887
+    },
+    {
+      "epoch": 0.8755237860488045,
+      "grad_norm": 0.16893395689368534,
+      "learning_rate": 8.055249706524575e-06,
+      "loss": 0.5199,
+      "step": 888
+    },
+    {
+      "epoch": 0.876509736258319,
+      "grad_norm": 0.16949385808256937,
+      "learning_rate": 8.051153802329963e-06,
+      "loss": 0.5016,
+      "step": 889
+    },
+    {
+      "epoch": 0.8774956864678334,
+      "grad_norm": 0.1927330298612067,
+      "learning_rate": 8.047054633202734e-06,
+      "loss": 0.5129,
+      "step": 890
+    },
+    {
+      "epoch": 0.8784816366773478,
+      "grad_norm": 0.17067432797447837,
+      "learning_rate": 8.042952203529262e-06,
+      "loss": 0.5326,
+      "step": 891
+    },
+    {
+      "epoch": 0.8794675868868622,
+      "grad_norm": 0.1739293602849575,
+      "learning_rate": 8.038846517699413e-06,
+      "loss": 0.5241,
+      "step": 892
+    },
+    {
+      "epoch": 0.8804535370963766,
+      "grad_norm": 0.171650640715769,
+      "learning_rate": 8.034737580106537e-06,
+      "loss": 0.5209,
+      "step": 893
+    },
+    {
+      "epoch": 0.8814394873058911,
+      "grad_norm": 0.1757326594314937,
+      "learning_rate": 8.030625395147467e-06,
+      "loss": 0.53,
+      "step": 894
+    },
+    {
+      "epoch": 0.8824254375154055,
+      "grad_norm": 0.1604081833605916,
+      "learning_rate": 8.026509967222504e-06,
+      "loss": 0.4722,
+      "step": 895
+    },
+    {
+      "epoch": 0.8834113877249199,
+      "grad_norm": 0.17236074876862664,
+      "learning_rate": 8.022391300735424e-06,
+      "loss": 0.5274,
+      "step": 896
+    },
+    {
+      "epoch": 0.8843973379344343,
+      "grad_norm": 0.1794389453398671,
+      "learning_rate": 8.01826940009347e-06,
+      "loss": 0.5235,
+      "step": 897
+    },
+    {
+      "epoch": 0.8853832881439487,
+      "grad_norm": 0.17308007813999723,
+      "learning_rate": 8.01414426970734e-06,
+      "loss": 0.5361,
+      "step": 898
+    },
+    {
+      "epoch": 0.8863692383534632,
+      "grad_norm": 0.18746313776847923,
+      "learning_rate": 8.010015913991194e-06,
+      "loss": 0.5124,
+      "step": 899
+    },
+    {
+      "epoch": 0.8873551885629776,
+      "grad_norm": 0.16596205512741172,
+      "learning_rate": 8.005884337362637e-06,
+      "loss": 0.5024,
+      "step": 900
+    },
+    {
+      "epoch": 0.888341138772492,
+      "grad_norm": 0.17098191624404319,
+      "learning_rate": 8.001749544242728e-06,
+      "loss": 0.5285,
+      "step": 901
+    },
+    {
+      "epoch": 0.8893270889820064,
+      "grad_norm": 0.17678419257324562,
+      "learning_rate": 7.997611539055962e-06,
+      "loss": 0.5244,
+      "step": 902
+    },
+    {
+      "epoch": 0.8903130391915208,
+      "grad_norm": 0.16721705873096537,
+      "learning_rate": 7.993470326230274e-06,
+      "loss": 0.5296,
+      "step": 903
+    },
+    {
+      "epoch": 0.8912989894010352,
+      "grad_norm": 0.17021479193248848,
+      "learning_rate": 7.98932591019703e-06,
+      "loss": 0.5162,
+      "step": 904
+    },
+    {
+      "epoch": 0.8922849396105497,
+      "grad_norm": 0.17301667833486167,
+      "learning_rate": 7.985178295391023e-06,
+      "loss": 0.5302,
+      "step": 905
+    },
+    {
+      "epoch": 0.8932708898200641,
+      "grad_norm": 0.1736113746385346,
+      "learning_rate": 7.981027486250472e-06,
+      "loss": 0.5193,
+      "step": 906
+    },
+    {
+      "epoch": 0.8942568400295785,
+      "grad_norm": 0.16199362086063365,
+      "learning_rate": 7.976873487217011e-06,
+      "loss": 0.5048,
+      "step": 907
+    },
+    {
+      "epoch": 0.8952427902390929,
+      "grad_norm": 0.17212503754715186,
+      "learning_rate": 7.972716302735692e-06,
+      "loss": 0.5225,
+      "step": 908
+    },
+    {
+      "epoch": 0.8962287404486073,
+      "grad_norm": 0.1645622560183533,
+      "learning_rate": 7.968555937254967e-06,
+      "loss": 0.5165,
+      "step": 909
+    },
+    {
+      "epoch": 0.8972146906581218,
+      "grad_norm": 0.16996160353433545,
+      "learning_rate": 7.964392395226699e-06,
+      "loss": 0.518,
+      "step": 910
+    },
+    {
+      "epoch": 0.8982006408676362,
+      "grad_norm": 0.17572896215315817,
+      "learning_rate": 7.960225681106151e-06,
+      "loss": 0.5247,
+      "step": 911
+    },
+    {
+      "epoch": 0.8991865910771506,
+      "grad_norm": 0.1719561893313999,
+      "learning_rate": 7.956055799351972e-06,
+      "loss": 0.5073,
+      "step": 912
+    },
+    {
+      "epoch": 0.900172541286665,
+      "grad_norm": 0.16090581754225922,
+      "learning_rate": 7.951882754426212e-06,
+      "loss": 0.516,
+      "step": 913
+    },
+    {
+      "epoch": 0.9011584914961794,
+      "grad_norm": 0.1807242712673357,
+      "learning_rate": 7.947706550794297e-06,
+      "loss": 0.5059,
+      "step": 914
+    },
+    {
+      "epoch": 0.9021444417056939,
+      "grad_norm": 0.17762020465728326,
+      "learning_rate": 7.943527192925035e-06,
+      "loss": 0.5195,
+      "step": 915
+    },
+    {
+      "epoch": 0.9031303919152083,
+      "grad_norm": 0.17141091502345296,
+      "learning_rate": 7.939344685290612e-06,
+      "loss": 0.5204,
+      "step": 916
+    },
+    {
+      "epoch": 0.9041163421247227,
+      "grad_norm": 0.17266717729532846,
+      "learning_rate": 7.935159032366583e-06,
+      "loss": 0.5046,
+      "step": 917
+    },
+    {
+      "epoch": 0.9051022923342371,
+      "grad_norm": 0.16644553924878852,
+      "learning_rate": 7.930970238631867e-06,
+      "loss": 0.5212,
+      "step": 918
+    },
+    {
+      "epoch": 0.9060882425437515,
+      "grad_norm": 0.1792204259622754,
+      "learning_rate": 7.926778308568746e-06,
+      "loss": 0.5103,
+      "step": 919
+    },
+    {
+      "epoch": 0.907074192753266,
+      "grad_norm": 0.16523094967902968,
+      "learning_rate": 7.922583246662858e-06,
+      "loss": 0.5136,
+      "step": 920
+    },
+    {
+      "epoch": 0.9080601429627804,
+      "grad_norm": 0.2119077911673333,
+      "learning_rate": 7.918385057403188e-06,
+      "loss": 0.5039,
+      "step": 921
+    },
+    {
+      "epoch": 0.9090460931722948,
+      "grad_norm": 0.17711635746479226,
+      "learning_rate": 7.914183745282076e-06,
+      "loss": 0.5233,
+      "step": 922
+    },
+    {
+      "epoch": 0.9100320433818092,
+      "grad_norm": 0.1660947936258377,
+      "learning_rate": 7.909979314795195e-06,
+      "loss": 0.5132,
+      "step": 923
+    },
+    {
+      "epoch": 0.9110179935913236,
+      "grad_norm": 0.16855795483064662,
+      "learning_rate": 7.905771770441559e-06,
+      "loss": 0.5065,
+      "step": 924
+    },
+    {
+      "epoch": 0.912003943800838,
+      "grad_norm": 0.16027620736877113,
+      "learning_rate": 7.901561116723518e-06,
+      "loss": 0.5091,
+      "step": 925
+    },
+    {
+      "epoch": 0.9129898940103525,
+      "grad_norm": 0.16175407780430773,
+      "learning_rate": 7.897347358146736e-06,
+      "loss": 0.5193,
+      "step": 926
+    },
+    {
+      "epoch": 0.9139758442198669,
+      "grad_norm": 0.16488932476570586,
+      "learning_rate": 7.893130499220216e-06,
+      "loss": 0.5123,
+      "step": 927
+    },
+    {
+      "epoch": 0.9149617944293813,
+      "grad_norm": 0.1734873791504118,
+      "learning_rate": 7.888910544456269e-06,
+      "loss": 0.5063,
+      "step": 928
+    },
+    {
+      "epoch": 0.9159477446388957,
+      "grad_norm": 0.17958456178996576,
+      "learning_rate": 7.884687498370519e-06,
+      "loss": 0.4909,
+      "step": 929
+    },
+    {
+      "epoch": 0.9169336948484101,
+      "grad_norm": 0.16643703344223132,
+      "learning_rate": 7.880461365481898e-06,
+      "loss": 0.5182,
+      "step": 930
+    },
+    {
+      "epoch": 0.9179196450579246,
+      "grad_norm": 0.1720747470600968,
+      "learning_rate": 7.876232150312646e-06,
+      "loss": 0.5151,
+      "step": 931
+    },
+    {
+      "epoch": 0.918905595267439,
+      "grad_norm": 0.17776188364898196,
+      "learning_rate": 7.871999857388295e-06,
+      "loss": 0.5183,
+      "step": 932
+    },
+    {
+      "epoch": 0.9198915454769534,
+      "grad_norm": 0.1615927283246447,
+      "learning_rate": 7.867764491237675e-06,
+      "loss": 0.5327,
+      "step": 933
+    },
+    {
+      "epoch": 0.9208774956864678,
+      "grad_norm": 0.1677229785390601,
+      "learning_rate": 7.863526056392904e-06,
+      "loss": 0.5269,
+      "step": 934
+    },
+    {
+      "epoch": 0.9218634458959822,
+      "grad_norm": 0.18531257466086812,
+      "learning_rate": 7.85928455738938e-06,
+      "loss": 0.5076,
+      "step": 935
+    },
+    {
+      "epoch": 0.9228493961054967,
+      "grad_norm": 0.17682682721854523,
+      "learning_rate": 7.855039998765781e-06,
+      "loss": 0.5311,
+      "step": 936
+    },
+    {
+      "epoch": 0.9238353463150111,
+      "grad_norm": 0.1595345354564208,
+      "learning_rate": 7.850792385064064e-06,
+      "loss": 0.5139,
+      "step": 937
+    },
+    {
+      "epoch": 0.9248212965245255,
+      "grad_norm": 0.1621031020930392,
+      "learning_rate": 7.846541720829448e-06,
+      "loss": 0.5065,
+      "step": 938
+    },
+    {
+      "epoch": 0.9258072467340399,
+      "grad_norm": 0.16982048292608531,
+      "learning_rate": 7.84228801061042e-06,
+      "loss": 0.5244,
+      "step": 939
+    },
+    {
+      "epoch": 0.9267931969435543,
+      "grad_norm": 0.164667046714563,
+      "learning_rate": 7.83803125895873e-06,
+      "loss": 0.4993,
+      "step": 940
+    },
+    {
+      "epoch": 0.9277791471530688,
+      "grad_norm": 0.1664120555152627,
+      "learning_rate": 7.833771470429375e-06,
+      "loss": 0.5236,
+      "step": 941
+    },
+    {
+      "epoch": 0.9287650973625832,
+      "grad_norm": 0.16363510468965828,
+      "learning_rate": 7.829508649580604e-06,
+      "loss": 0.483,
+      "step": 942
+    },
+    {
+      "epoch": 0.9297510475720976,
+      "grad_norm": 0.18300991114328144,
+      "learning_rate": 7.825242800973915e-06,
+      "loss": 0.5052,
+      "step": 943
+    },
+    {
+      "epoch": 0.930736997781612,
+      "grad_norm": 0.16805069209354068,
+      "learning_rate": 7.82097392917404e-06,
+      "loss": 0.5023,
+      "step": 944
+    },
+    {
+      "epoch": 0.9317229479911264,
+      "grad_norm": 0.17945312719468895,
+      "learning_rate": 7.816702038748953e-06,
+      "loss": 0.5206,
+      "step": 945
+    },
+    {
+      "epoch": 0.9327088982006408,
+      "grad_norm": 0.16520942441751493,
+      "learning_rate": 7.812427134269852e-06,
+      "loss": 0.5054,
+      "step": 946
+    },
+    {
+      "epoch": 0.9336948484101553,
+      "grad_norm": 0.18163830376615256,
+      "learning_rate": 7.80814922031116e-06,
+      "loss": 0.5371,
+      "step": 947
+    },
+    {
+      "epoch": 0.9346807986196697,
+      "grad_norm": 0.16622720857375242,
+      "learning_rate": 7.803868301450528e-06,
+      "loss": 0.5063,
+      "step": 948
+    },
+    {
+      "epoch": 0.9356667488291841,
+      "grad_norm": 0.17626958165949663,
+      "learning_rate": 7.79958438226881e-06,
+      "loss": 0.5154,
+      "step": 949
+    },
+    {
+      "epoch": 0.9366526990386985,
+      "grad_norm": 0.17179172076094043,
+      "learning_rate": 7.795297467350083e-06,
+      "loss": 0.5056,
+      "step": 950
+    },
+    {
+      "epoch": 0.9376386492482129,
+      "grad_norm": 0.17404831583610947,
+      "learning_rate": 7.791007561281623e-06,
+      "loss": 0.504,
+      "step": 951
+    },
+    {
+      "epoch": 0.9386245994577274,
+      "grad_norm": 0.17341440294441673,
+      "learning_rate": 7.786714668653907e-06,
+      "loss": 0.4975,
+      "step": 952
+    },
+    {
+      "epoch": 0.9396105496672418,
+      "grad_norm": 0.17106305253040335,
+      "learning_rate": 7.782418794060609e-06,
+      "loss": 0.5304,
+      "step": 953
+    },
+    {
+      "epoch": 0.9405964998767562,
+      "grad_norm": 0.1657714097764547,
+      "learning_rate": 7.778119942098594e-06,
+      "loss": 0.5008,
+      "step": 954
+    },
+    {
+      "epoch": 0.9415824500862706,
+      "grad_norm": 0.17215962243937322,
+      "learning_rate": 7.773818117367913e-06,
+      "loss": 0.5039,
+      "step": 955
+    },
+    {
+      "epoch": 0.942568400295785,
+      "grad_norm": 0.1708309325645627,
+      "learning_rate": 7.769513324471798e-06,
+      "loss": 0.4907,
+      "step": 956
+    },
+    {
+      "epoch": 0.9435543505052995,
+      "grad_norm": 0.16824173694397346,
+      "learning_rate": 7.765205568016654e-06,
+      "loss": 0.5279,
+      "step": 957
+    },
+    {
+      "epoch": 0.9445403007148139,
+      "grad_norm": 0.1631371599910111,
+      "learning_rate": 7.760894852612064e-06,
+      "loss": 0.5017,
+      "step": 958
+    },
+    {
+      "epoch": 0.9455262509243283,
+      "grad_norm": 0.1654800873573575,
+      "learning_rate": 7.75658118287077e-06,
+      "loss": 0.5011,
+      "step": 959
+    },
+    {
+      "epoch": 0.9465122011338427,
+      "grad_norm": 0.1701294258575521,
+      "learning_rate": 7.75226456340868e-06,
+      "loss": 0.5028,
+      "step": 960
+    },
+    {
+      "epoch": 0.9474981513433571,
+      "grad_norm": 0.17081358124511606,
+      "learning_rate": 7.747944998844858e-06,
+      "loss": 0.5147,
+      "step": 961
+    },
+    {
+      "epoch": 0.9484841015528716,
+      "grad_norm": 0.1702922338705157,
+      "learning_rate": 7.743622493801518e-06,
+      "loss": 0.5247,
+      "step": 962
+    },
+    {
+      "epoch": 0.949470051762386,
+      "grad_norm": 0.16354067278521262,
+      "learning_rate": 7.739297052904018e-06,
+      "loss": 0.508,
+      "step": 963
+    },
+    {
+      "epoch": 0.9504560019719004,
+      "grad_norm": 0.1632330387681569,
+      "learning_rate": 7.734968680780865e-06,
+      "loss": 0.5197,
+      "step": 964
+    },
+    {
+      "epoch": 0.9514419521814148,
+      "grad_norm": 0.16707232934194988,
+      "learning_rate": 7.730637382063696e-06,
+      "loss": 0.5168,
+      "step": 965
+    },
+    {
+      "epoch": 0.9524279023909292,
+      "grad_norm": 0.1698770968536073,
+      "learning_rate": 7.72630316138728e-06,
+      "loss": 0.501,
+      "step": 966
+    },
+    {
+      "epoch": 0.9534138526004436,
+      "grad_norm": 0.16574650052795398,
+      "learning_rate": 7.721966023389519e-06,
+      "loss": 0.481,
+      "step": 967
+    },
+    {
+      "epoch": 0.9543998028099581,
+      "grad_norm": 0.1765087940076975,
+      "learning_rate": 7.717625972711429e-06,
+      "loss": 0.5018,
+      "step": 968
+    },
+    {
+      "epoch": 0.9553857530194725,
+      "grad_norm": 0.16999695085713057,
+      "learning_rate": 7.713283013997145e-06,
+      "loss": 0.5117,
+      "step": 969
+    },
+    {
+      "epoch": 0.9563717032289869,
+      "grad_norm": 0.16492375715904378,
+      "learning_rate": 7.708937151893917e-06,
+      "loss": 0.5295,
+      "step": 970
+    },
+    {
+      "epoch": 0.9573576534385013,
+      "grad_norm": 0.1770079333459357,
+      "learning_rate": 7.704588391052099e-06,
+      "loss": 0.5096,
+      "step": 971
+    },
+    {
+      "epoch": 0.9583436036480157,
+      "grad_norm": 0.26749772764267893,
+      "learning_rate": 7.700236736125146e-06,
+      "loss": 0.5036,
+      "step": 972
+    },
+    {
+      "epoch": 0.9593295538575302,
+      "grad_norm": 0.16742320355306792,
+      "learning_rate": 7.695882191769614e-06,
+      "loss": 0.513,
+      "step": 973
+    },
+    {
+      "epoch": 0.9603155040670446,
+      "grad_norm": 0.1654819848687592,
+      "learning_rate": 7.691524762645147e-06,
+      "loss": 0.5019,
+      "step": 974
+    },
+    {
+      "epoch": 0.961301454276559,
+      "grad_norm": 0.17272585215388925,
+      "learning_rate": 7.687164453414475e-06,
+      "loss": 0.53,
+      "step": 975
+    },
+    {
+      "epoch": 0.9622874044860734,
+      "grad_norm": 0.17361850361878714,
+      "learning_rate": 7.682801268743413e-06,
+      "loss": 0.5216,
+      "step": 976
+    },
+    {
+      "epoch": 0.9632733546955878,
+      "grad_norm": 0.1705371849345429,
+      "learning_rate": 7.678435213300851e-06,
+      "loss": 0.5023,
+      "step": 977
+    },
+    {
+      "epoch": 0.9642593049051023,
+      "grad_norm": 0.16294961746913053,
+      "learning_rate": 7.674066291758756e-06,
+      "loss": 0.4913,
+      "step": 978
+    },
+    {
+      "epoch": 0.9652452551146167,
+      "grad_norm": 0.16631852074052206,
+      "learning_rate": 7.669694508792153e-06,
+      "loss": 0.5192,
+      "step": 979
+    },
+    {
+      "epoch": 0.9662312053241311,
+      "grad_norm": 0.1703772839195891,
+      "learning_rate": 7.665319869079136e-06,
+      "loss": 0.5094,
+      "step": 980
+    },
+    {
+      "epoch": 0.9672171555336455,
+      "grad_norm": 0.16104122981289215,
+      "learning_rate": 7.660942377300853e-06,
+      "loss": 0.5045,
+      "step": 981
+    },
+    {
+      "epoch": 0.9682031057431599,
+      "grad_norm": 0.1687999170074842,
+      "learning_rate": 7.656562038141502e-06,
+      "loss": 0.518,
+      "step": 982
+    },
+    {
+      "epoch": 0.9691890559526743,
+      "grad_norm": 0.1640142258647821,
+      "learning_rate": 7.652178856288333e-06,
+      "loss": 0.5123,
+      "step": 983
+    },
+    {
+      "epoch": 0.9701750061621888,
+      "grad_norm": 0.29857681452884277,
+      "learning_rate": 7.647792836431633e-06,
+      "loss": 0.5014,
+      "step": 984
+    },
+    {
+      "epoch": 0.9711609563717032,
+      "grad_norm": 0.166417158069653,
+      "learning_rate": 7.643403983264733e-06,
+      "loss": 0.5174,
+      "step": 985
+    },
+    {
+      "epoch": 0.9721469065812176,
+      "grad_norm": 0.17976511665766357,
+      "learning_rate": 7.639012301483983e-06,
+      "loss": 0.51,
+      "step": 986
+    },
+    {
+      "epoch": 0.973132856790732,
+      "grad_norm": 0.16710700103318898,
+      "learning_rate": 7.634617795788773e-06,
+      "loss": 0.508,
+      "step": 987
+    },
+    {
+      "epoch": 0.9741188070002464,
+      "grad_norm": 0.16652682426124865,
+      "learning_rate": 7.630220470881506e-06,
+      "loss": 0.4879,
+      "step": 988
+    },
+    {
+      "epoch": 0.9751047572097609,
+      "grad_norm": 0.18519334923187553,
+      "learning_rate": 7.6258203314676105e-06,
+      "loss": 0.523,
+      "step": 989
+    },
+    {
+      "epoch": 0.9760907074192753,
+      "grad_norm": 0.1769398076186771,
+      "learning_rate": 7.621417382255516e-06,
+      "loss": 0.4924,
+      "step": 990
+    },
+    {
+      "epoch": 0.9770766576287897,
+      "grad_norm": 0.1749755537077236,
+      "learning_rate": 7.617011627956665e-06,
+      "loss": 0.5261,
+      "step": 991
+    },
+    {
+      "epoch": 0.9780626078383041,
+      "grad_norm": 0.19223145523362156,
+      "learning_rate": 7.612603073285503e-06,
+      "loss": 0.5277,
+      "step": 992
+    },
+    {
+      "epoch": 0.9790485580478185,
+      "grad_norm": 0.16720986150018388,
+      "learning_rate": 7.608191722959466e-06,
+      "loss": 0.5331,
+      "step": 993
+    },
+    {
+      "epoch": 0.980034508257333,
+      "grad_norm": 0.17106542111837608,
+      "learning_rate": 7.6037775816989875e-06,
+      "loss": 0.5117,
+      "step": 994
+    },
+    {
+      "epoch": 0.9810204584668474,
+      "grad_norm": 0.17579588202811694,
+      "learning_rate": 7.599360654227485e-06,
+      "loss": 0.5151,
+      "step": 995
+    },
+    {
+      "epoch": 0.9820064086763618,
+      "grad_norm": 0.17106435785907712,
+      "learning_rate": 7.5949409452713585e-06,
+      "loss": 0.5201,
+      "step": 996
+    },
+    {
+      "epoch": 0.9829923588858762,
+      "grad_norm": 0.16347983018754825,
+      "learning_rate": 7.590518459559981e-06,
+      "loss": 0.5084,
+      "step": 997
+    },
+    {
+      "epoch": 0.9839783090953906,
+      "grad_norm": 0.18181892193197213,
+      "learning_rate": 7.586093201825702e-06,
+      "loss": 0.5467,
+      "step": 998
+    },
+    {
+      "epoch": 0.984964259304905,
+      "grad_norm": 0.16737675009795525,
+      "learning_rate": 7.581665176803832e-06,
+      "loss": 0.5264,
+      "step": 999
+    },
+    {
+      "epoch": 0.9859502095144195,
+      "grad_norm": 0.1698415867558201,
+      "learning_rate": 7.577234389232646e-06,
+      "loss": 0.5005,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9869361597239339,
+      "grad_norm": 0.17199239759519194,
+      "learning_rate": 7.572800843853376e-06,
+      "loss": 0.5128,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9879221099334483,
+      "grad_norm": 0.18133702646379646,
+      "learning_rate": 7.568364545410201e-06,
+      "loss": 0.5225,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9889080601429627,
+      "grad_norm": 0.19334275561566608,
+      "learning_rate": 7.563925498650248e-06,
+      "loss": 0.5261,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9898940103524771,
+      "grad_norm": 0.17381612732882007,
+      "learning_rate": 7.5594837083235894e-06,
+      "loss": 0.5072,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9908799605619916,
+      "grad_norm": 0.16380698462662358,
+      "learning_rate": 7.555039179183223e-06,
+      "loss": 0.5049,
+      "step": 1005
+    },
+    {
+      "epoch": 0.991865910771506,
+      "grad_norm": 0.17462198651108224,
+      "learning_rate": 7.55059191598509e-06,
+      "loss": 0.5277,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9928518609810204,
+      "grad_norm": 0.17953006072611008,
+      "learning_rate": 7.546141923488045e-06,
+      "loss": 0.5186,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9938378111905348,
+      "grad_norm": 0.16865358332826616,
+      "learning_rate": 7.541689206453873e-06,
+      "loss": 0.5244,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9948237614000492,
+      "grad_norm": 0.17304916526129824,
+      "learning_rate": 7.5372337696472674e-06,
+      "loss": 0.5171,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9958097116095637,
+      "grad_norm": 0.17637586538791555,
+      "learning_rate": 7.532775617835836e-06,
+      "loss": 0.5249,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9967956618190781,
+      "grad_norm": 0.16779739755098547,
+      "learning_rate": 7.528314755790089e-06,
+      "loss": 0.5117,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9977816120285925,
+      "grad_norm": 0.16941643606585818,
+      "learning_rate": 7.523851188283442e-06,
+      "loss": 0.4903,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9987675622381069,
+      "grad_norm": 0.2093863707411628,
+      "learning_rate": 7.5193849200921986e-06,
+      "loss": 0.537,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9997535124476213,
+      "grad_norm": 0.16661642473837618,
+      "learning_rate": 7.514915955995558e-06,
+      "loss": 0.5032,
+      "step": 1014
+    },
+    {
+      "epoch": 1.0007394626571358,
+      "grad_norm": 0.17389943690378837,
+      "learning_rate": 7.510444300775599e-06,
+      "loss": 0.524,
+      "step": 1015
+    },
+    {
+      "epoch": 1.0017254128666502,
+      "grad_norm": 0.17311914260770017,
+      "learning_rate": 7.505969959217285e-06,
+      "loss": 0.5028,
+      "step": 1016
+    },
+    {
+      "epoch": 1.0029578506285433,
+      "grad_norm": 0.19178452526435935,
+      "learning_rate": 7.501492936108454e-06,
+      "loss": 0.4834,
+      "step": 1017
+    },
+    {
+      "epoch": 1.0039438008380577,
+      "grad_norm": 0.16819127155730185,
+      "learning_rate": 7.497013236239805e-06,
+      "loss": 0.4673,
+      "step": 1018
+    },
+    {
+      "epoch": 1.004929751047572,
+      "grad_norm": 0.2087088082599385,
+      "learning_rate": 7.492530864404916e-06,
+      "loss": 0.463,
+      "step": 1019
+    },
+    {
+      "epoch": 1.0059157012570865,
+      "grad_norm": 0.17023882357913672,
+      "learning_rate": 7.488045825400208e-06,
+      "loss": 0.465,
+      "step": 1020
+    },
+    {
+      "epoch": 1.006901651466601,
+      "grad_norm": 0.19096706186861023,
+      "learning_rate": 7.483558124024968e-06,
+      "loss": 0.4836,
+      "step": 1021
+    },
+    {
+      "epoch": 1.0078876016761154,
+      "grad_norm": 0.17292587142561078,
+      "learning_rate": 7.479067765081327e-06,
+      "loss": 0.4647,
+      "step": 1022
+    },
+    {
+      "epoch": 1.0088735518856298,
+      "grad_norm": 0.20970851321219616,
+      "learning_rate": 7.4745747533742604e-06,
+      "loss": 0.481,
+      "step": 1023
+    },
+    {
+      "epoch": 1.0098595020951442,
+      "grad_norm": 0.19957414505514495,
+      "learning_rate": 7.470079093711583e-06,
+      "loss": 0.4819,
+      "step": 1024
+    },
+    {
+      "epoch": 1.0108454523046586,
+      "grad_norm": 0.17235557104698107,
+      "learning_rate": 7.465580790903941e-06,
+      "loss": 0.4664,
+      "step": 1025
+    },
+    {
+      "epoch": 1.011831402514173,
+      "grad_norm": 0.17099614783137973,
+      "learning_rate": 7.461079849764812e-06,
+      "loss": 0.4594,
+      "step": 1026
+    },
+    {
+      "epoch": 1.0128173527236874,
+      "grad_norm": 0.17371113860745846,
+      "learning_rate": 7.456576275110495e-06,
+      "loss": 0.4627,
+      "step": 1027
+    },
+    {
+      "epoch": 1.0138033029332019,
+      "grad_norm": 0.18992048708241716,
+      "learning_rate": 7.452070071760106e-06,
+      "loss": 0.4753,
+      "step": 1028
+    },
+    {
+      "epoch": 1.0147892531427163,
+      "grad_norm": 0.17058713785573343,
+      "learning_rate": 7.447561244535575e-06,
+      "loss": 0.4693,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0157752033522307,
+      "grad_norm": 0.1696152852494023,
+      "learning_rate": 7.443049798261643e-06,
+      "loss": 0.4748,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0167611535617451,
+      "grad_norm": 0.17311444766409678,
+      "learning_rate": 7.438535737765846e-06,
+      "loss": 0.4738,
+      "step": 1031
+    },
+    {
+      "epoch": 1.0177471037712595,
+      "grad_norm": 0.1789380413927639,
+      "learning_rate": 7.434019067878524e-06,
+      "loss": 0.4868,
+      "step": 1032
+    },
+    {
+      "epoch": 1.018733053980774,
+      "grad_norm": 0.1776370707448934,
+      "learning_rate": 7.429499793432806e-06,
+      "loss": 0.4639,
+      "step": 1033
+    },
+    {
+      "epoch": 1.0197190041902884,
+      "grad_norm": 0.16321989041138424,
+      "learning_rate": 7.424977919264611e-06,
+      "loss": 0.4646,
+      "step": 1034
+    },
+    {
+      "epoch": 1.0207049543998028,
+      "grad_norm": 0.3565072795267459,
+      "learning_rate": 7.420453450212635e-06,
+      "loss": 0.4731,
+      "step": 1035
+    },
+    {
+      "epoch": 1.0216909046093172,
+      "grad_norm": 0.1718732951805992,
+      "learning_rate": 7.415926391118357e-06,
+      "loss": 0.4734,
+      "step": 1036
+    },
+    {
+      "epoch": 1.0226768548188316,
+      "grad_norm": 0.17702243568254172,
+      "learning_rate": 7.41139674682602e-06,
+      "loss": 0.4752,
+      "step": 1037
+    },
+    {
+      "epoch": 1.023662805028346,
+      "grad_norm": 0.1802015150153324,
+      "learning_rate": 7.4068645221826415e-06,
+      "loss": 0.467,
+      "step": 1038
+    },
+    {
+      "epoch": 1.0246487552378605,
+      "grad_norm": 0.17785257353690725,
+      "learning_rate": 7.402329722037993e-06,
+      "loss": 0.464,
+      "step": 1039
+    },
+    {
+      "epoch": 1.025634705447375,
+      "grad_norm": 0.17670639846700095,
+      "learning_rate": 7.397792351244607e-06,
+      "loss": 0.461,
+      "step": 1040
+    },
+    {
+      "epoch": 1.0266206556568893,
+      "grad_norm": 0.16971386522429677,
+      "learning_rate": 7.393252414657762e-06,
+      "loss": 0.4915,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0276066058664037,
+      "grad_norm": 0.21567966408791525,
+      "learning_rate": 7.388709917135489e-06,
+      "loss": 0.4647,
+      "step": 1042
+    },
+    {
+      "epoch": 1.0285925560759182,
+      "grad_norm": 0.171778494469584,
+      "learning_rate": 7.3841648635385525e-06,
+      "loss": 0.4772,
+      "step": 1043
+    },
+    {
+      "epoch": 1.0295785062854326,
+      "grad_norm": 0.17106503306799054,
+      "learning_rate": 7.379617258730456e-06,
+      "loss": 0.4684,
+      "step": 1044
+    },
+    {
+      "epoch": 1.030564456494947,
+      "grad_norm": 0.17303586751910072,
+      "learning_rate": 7.375067107577428e-06,
+      "loss": 0.4829,
+      "step": 1045
+    },
+    {
+      "epoch": 1.0315504067044614,
+      "grad_norm": 0.18405366877277182,
+      "learning_rate": 7.370514414948432e-06,
+      "loss": 0.4874,
+      "step": 1046
+    },
+    {
+      "epoch": 1.0325363569139758,
+      "grad_norm": 0.1737527019460879,
+      "learning_rate": 7.3659591857151405e-06,
+      "loss": 0.4643,
+      "step": 1047
+    },
+    {
+      "epoch": 1.0335223071234902,
+      "grad_norm": 0.38823443343321035,
+      "learning_rate": 7.361401424751945e-06,
+      "loss": 0.47,
+      "step": 1048
+    },
+    {
+      "epoch": 1.0345082573330047,
+      "grad_norm": 0.17985440451371562,
+      "learning_rate": 7.356841136935946e-06,
+      "loss": 0.4926,
+      "step": 1049
+    },
+    {
+      "epoch": 1.035494207542519,
+      "grad_norm": 0.1655664487500876,
+      "learning_rate": 7.352278327146946e-06,
+      "loss": 0.4631,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0364801577520335,
+      "grad_norm": 0.16699537292326075,
+      "learning_rate": 7.347713000267451e-06,
+      "loss": 0.4775,
+      "step": 1051
+    },
+    {
+      "epoch": 1.037466107961548,
+      "grad_norm": 0.1653107310751375,
+      "learning_rate": 7.343145161182654e-06,
+      "loss": 0.4486,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0384520581710623,
+      "grad_norm": 0.20928144693777923,
+      "learning_rate": 7.338574814780442e-06,
+      "loss": 0.4714,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0394380083805768,
+      "grad_norm": 0.16832051275492646,
+      "learning_rate": 7.33400196595138e-06,
+      "loss": 0.4689,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0404239585900912,
+      "grad_norm": 0.17097271806660913,
+      "learning_rate": 7.329426619588713e-06,
+      "loss": 0.4737,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0414099087996056,
+      "grad_norm": 0.16375357742355764,
+      "learning_rate": 7.324848780588359e-06,
+      "loss": 0.464,
+      "step": 1056
+    },
+    {
+      "epoch": 1.04239585900912,
+      "grad_norm": 0.16418865158498658,
+      "learning_rate": 7.3202684538489056e-06,
+      "loss": 0.4629,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0433818092186344,
+      "grad_norm": 0.16611259691226854,
+      "learning_rate": 7.315685644271595e-06,
+      "loss": 0.4667,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0443677594281489,
+      "grad_norm": 0.17033513470900657,
+      "learning_rate": 7.311100356760334e-06,
+      "loss": 0.473,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0453537096376633,
+      "grad_norm": 0.17518713238225492,
+      "learning_rate": 7.306512596221678e-06,
+      "loss": 0.4884,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0463396598471777,
+      "grad_norm": 0.20198021631299554,
+      "learning_rate": 7.301922367564828e-06,
+      "loss": 0.4778,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0473256100566921,
+      "grad_norm": 0.16412321917522685,
+      "learning_rate": 7.297329675701625e-06,
+      "loss": 0.462,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0483115602662065,
+      "grad_norm": 0.17500314542240664,
+      "learning_rate": 7.29273452554655e-06,
+      "loss": 0.4607,
+      "step": 1063
+    },
+    {
+      "epoch": 1.049297510475721,
+      "grad_norm": 0.1662387855658783,
+      "learning_rate": 7.28813692201671e-06,
+      "loss": 0.4722,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0502834606852354,
+      "grad_norm": 0.1651013643582452,
+      "learning_rate": 7.283536870031841e-06,
+      "loss": 0.455,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0512694108947498,
+      "grad_norm": 0.1980382070465155,
+      "learning_rate": 7.278934374514295e-06,
+      "loss": 0.4727,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0522553611042642,
+      "grad_norm": 0.21478207995676313,
+      "learning_rate": 7.274329440389043e-06,
+      "loss": 0.4817,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0532413113137786,
+      "grad_norm": 0.1669942279540382,
+      "learning_rate": 7.269722072583661e-06,
+      "loss": 0.4667,
+      "step": 1068
+    },
+    {
+      "epoch": 1.054227261523293,
+      "grad_norm": 0.1833073183115231,
+      "learning_rate": 7.265112276028334e-06,
+      "loss": 0.4729,
+      "step": 1069
+    },
+    {
+      "epoch": 1.0552132117328075,
+      "grad_norm": 0.16101246541844533,
+      "learning_rate": 7.260500055655843e-06,
+      "loss": 0.4605,
+      "step": 1070
+    },
+    {
+      "epoch": 1.0561991619423219,
+      "grad_norm": 0.1724004172085386,
+      "learning_rate": 7.255885416401565e-06,
+      "loss": 0.4557,
+      "step": 1071
+    },
+    {
+      "epoch": 1.0571851121518363,
+      "grad_norm": 0.17652668859621282,
+      "learning_rate": 7.251268363203458e-06,
+      "loss": 0.4679,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0581710623613507,
+      "grad_norm": 0.16265501174831257,
+      "learning_rate": 7.246648901002073e-06,
+      "loss": 0.4623,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0591570125708651,
+      "grad_norm": 0.1681511080215305,
+      "learning_rate": 7.242027034740533e-06,
+      "loss": 0.4741,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0601429627803796,
+      "grad_norm": 0.16691945200969913,
+      "learning_rate": 7.2374027693645364e-06,
+      "loss": 0.4738,
+      "step": 1075
+    },
+    {
+      "epoch": 1.061128912989894,
+      "grad_norm": 0.1753163266766736,
+      "learning_rate": 7.232776109822346e-06,
+      "loss": 0.4749,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0621148631994084,
+      "grad_norm": 0.16592219858586765,
+      "learning_rate": 7.2281470610647885e-06,
+      "loss": 0.476,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0631008134089228,
+      "grad_norm": 0.168087769220742,
+      "learning_rate": 7.223515628045246e-06,
+      "loss": 0.4617,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0640867636184372,
+      "grad_norm": 0.3051422068414839,
+      "learning_rate": 7.218881815719651e-06,
+      "loss": 0.4691,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0650727138279517,
+      "grad_norm": 0.16389945371975692,
+      "learning_rate": 7.214245629046488e-06,
+      "loss": 0.4565,
+      "step": 1080
+    },
+    {
+      "epoch": 1.066058664037466,
+      "grad_norm": 0.1646261439122491,
+      "learning_rate": 7.209607072986772e-06,
+      "loss": 0.4889,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0670446142469805,
+      "grad_norm": 0.5052895776621189,
+      "learning_rate": 7.204966152504064e-06,
+      "loss": 0.4718,
+      "step": 1082
+    },
+    {
+      "epoch": 1.068030564456495,
+      "grad_norm": 0.16767373688606013,
+      "learning_rate": 7.200322872564444e-06,
+      "loss": 0.4503,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0690165146660093,
+      "grad_norm": 0.17344310017547962,
+      "learning_rate": 7.195677238136532e-06,
+      "loss": 0.4584,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0700024648755238,
+      "grad_norm": 0.17012418880271177,
+      "learning_rate": 7.1910292541914505e-06,
+      "loss": 0.4901,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0709884150850382,
+      "grad_norm": 0.17520849654487555,
+      "learning_rate": 7.186378925702847e-06,
+      "loss": 0.4632,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0719743652945526,
+      "grad_norm": 0.16694682563355986,
+      "learning_rate": 7.181726257646875e-06,
+      "loss": 0.4632,
+      "step": 1087
+    },
+    {
+      "epoch": 1.072960315504067,
+      "grad_norm": 0.16679362368659548,
+      "learning_rate": 7.17707125500219e-06,
+      "loss": 0.4531,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0739462657135814,
+      "grad_norm": 0.17233737738544838,
+      "learning_rate": 7.172413922749949e-06,
+      "loss": 0.4681,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0749322159230958,
+      "grad_norm": 0.17216500444653932,
+      "learning_rate": 7.167754265873799e-06,
+      "loss": 0.457,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0759181661326103,
+      "grad_norm": 0.1717603157377839,
+      "learning_rate": 7.163092289359874e-06,
+      "loss": 0.4582,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0769041163421247,
+      "grad_norm": 0.17624404758316678,
+      "learning_rate": 7.158427998196794e-06,
+      "loss": 0.4498,
+      "step": 1092
+    },
+    {
+      "epoch": 1.077890066551639,
+      "grad_norm": 0.1643558553861484,
+      "learning_rate": 7.15376139737565e-06,
+      "loss": 0.467,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0788760167611535,
+      "grad_norm": 0.16438382664576945,
+      "learning_rate": 7.149092491890012e-06,
+      "loss": 0.4725,
+      "step": 1094
+    },
+    {
+      "epoch": 1.079861966970668,
+      "grad_norm": 0.16037568324882467,
+      "learning_rate": 7.144421286735907e-06,
+      "loss": 0.4897,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0808479171801824,
+      "grad_norm": 0.1736892235376666,
+      "learning_rate": 7.139747786911833e-06,
+      "loss": 0.4555,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0818338673896968,
+      "grad_norm": 0.16573370625810585,
+      "learning_rate": 7.135071997418733e-06,
+      "loss": 0.4825,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0828198175992112,
+      "grad_norm": 0.17245174768386456,
+      "learning_rate": 7.130393923260008e-06,
+      "loss": 0.4726,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0838057678087256,
+      "grad_norm": 0.16248610358495352,
+      "learning_rate": 7.125713569441502e-06,
+      "loss": 0.457,
+      "step": 1099
+    },
+    {
+      "epoch": 1.08479171801824,
+      "grad_norm": 0.1645857835183841,
+      "learning_rate": 7.121030940971496e-06,
+      "loss": 0.471,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0857776682277545,
+      "grad_norm": 0.17570847262336248,
+      "learning_rate": 7.1163460428607065e-06,
+      "loss": 0.4836,
+      "step": 1101
+    },
+    {
+      "epoch": 1.0867636184372689,
+      "grad_norm": 0.16498768564272298,
+      "learning_rate": 7.1116588801222785e-06,
+      "loss": 0.4694,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0877495686467833,
+      "grad_norm": 0.16798012630933712,
+      "learning_rate": 7.106969457771782e-06,
+      "loss": 0.465,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0887355188562977,
+      "grad_norm": 0.224016566078945,
+      "learning_rate": 7.102277780827198e-06,
+      "loss": 0.4653,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0897214690658121,
+      "grad_norm": 0.17016452650065036,
+      "learning_rate": 7.097583854308934e-06,
+      "loss": 0.4875,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0907074192753266,
+      "grad_norm": 0.16271342027732072,
+      "learning_rate": 7.092887683239786e-06,
+      "loss": 0.4696,
+      "step": 1106
+    },
+    {
+      "epoch": 1.091693369484841,
+      "grad_norm": 0.16688527743939172,
+      "learning_rate": 7.088189272644971e-06,
+      "loss": 0.4622,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0926793196943554,
+      "grad_norm": 0.16638349989284135,
+      "learning_rate": 7.083488627552089e-06,
+      "loss": 0.4709,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0936652699038698,
+      "grad_norm": 0.16461375486233376,
+      "learning_rate": 7.078785752991134e-06,
+      "loss": 0.4714,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0946512201133842,
+      "grad_norm": 0.17605419735885408,
+      "learning_rate": 7.074080653994491e-06,
+      "loss": 0.469,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0956371703228986,
+      "grad_norm": 0.16324910009751914,
+      "learning_rate": 7.069373335596918e-06,
+      "loss": 0.466,
+      "step": 1111
+    },
+    {
+      "epoch": 1.096623120532413,
+      "grad_norm": 0.2943390226786043,
+      "learning_rate": 7.0646638028355515e-06,
+      "loss": 0.4672,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0976090707419275,
+      "grad_norm": 0.16732350138152158,
+      "learning_rate": 7.0599520607499e-06,
+      "loss": 0.4764,
+      "step": 1113
+    },
+    {
+      "epoch": 1.098595020951442,
+      "grad_norm": 0.1692670065093426,
+      "learning_rate": 7.0552381143818295e-06,
+      "loss": 0.4645,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0995809711609563,
+      "grad_norm": 0.17387257746846765,
+      "learning_rate": 7.050521968775574e-06,
+      "loss": 0.4751,
+      "step": 1115
+    },
+    {
+      "epoch": 1.1005669213704707,
+      "grad_norm": 0.16017857772958252,
+      "learning_rate": 7.045803628977708e-06,
+      "loss": 0.473,
+      "step": 1116
+    },
+    {
+      "epoch": 1.1015528715799852,
+      "grad_norm": 0.16293349123194426,
+      "learning_rate": 7.041083100037167e-06,
+      "loss": 0.4607,
+      "step": 1117
+    },
+    {
+      "epoch": 1.1025388217894996,
+      "grad_norm": 0.17002810017550496,
+      "learning_rate": 7.036360387005223e-06,
+      "loss": 0.4827,
+      "step": 1118
+    },
+    {
+      "epoch": 1.103524771999014,
+      "grad_norm": 0.15772025383880783,
+      "learning_rate": 7.031635494935483e-06,
+      "loss": 0.4478,
+      "step": 1119
+    },
+    {
+      "epoch": 1.1045107222085284,
+      "grad_norm": 0.16545767237670908,
+      "learning_rate": 7.02690842888389e-06,
+      "loss": 0.4746,
+      "step": 1120
+    },
+    {
+      "epoch": 1.1054966724180428,
+      "grad_norm": 0.16915171076158128,
+      "learning_rate": 7.02217919390871e-06,
+      "loss": 0.4788,
+      "step": 1121
+    },
+    {
+      "epoch": 1.1064826226275573,
+      "grad_norm": 0.168632485475845,
+      "learning_rate": 7.017447795070533e-06,
+      "loss": 0.4846,
+      "step": 1122
+    },
+    {
+      "epoch": 1.1074685728370717,
+      "grad_norm": 0.16577222327335642,
+      "learning_rate": 7.0127142374322634e-06,
+      "loss": 0.4624,
+      "step": 1123
+    },
+    {
+      "epoch": 1.108454523046586,
+      "grad_norm": 0.16593246110678658,
+      "learning_rate": 7.007978526059113e-06,
+      "loss": 0.4646,
+      "step": 1124
+    },
+    {
+      "epoch": 1.1094404732561005,
+      "grad_norm": 0.1622469903471737,
+      "learning_rate": 7.003240666018602e-06,
+      "loss": 0.4593,
+      "step": 1125
+    },
+    {
+      "epoch": 1.110426423465615,
+      "grad_norm": 0.16740371758851988,
+      "learning_rate": 6.998500662380547e-06,
+      "loss": 0.4769,
+      "step": 1126
+    },
+    {
+      "epoch": 1.1114123736751294,
+      "grad_norm": 0.16729177647362864,
+      "learning_rate": 6.993758520217059e-06,
+      "loss": 0.4746,
+      "step": 1127
+    },
+    {
+      "epoch": 1.1123983238846438,
+      "grad_norm": 0.17152389487000505,
+      "learning_rate": 6.989014244602541e-06,
+      "loss": 0.4617,
+      "step": 1128
+    },
+    {
+      "epoch": 1.1133842740941582,
+      "grad_norm": 0.16258445009175337,
+      "learning_rate": 6.984267840613672e-06,
+      "loss": 0.4769,
+      "step": 1129
+    },
+    {
+      "epoch": 1.1143702243036726,
+      "grad_norm": 0.17440742675904922,
+      "learning_rate": 6.979519313329417e-06,
+      "loss": 0.4731,
+      "step": 1130
+    },
+    {
+      "epoch": 1.115356174513187,
+      "grad_norm": 0.16660447566325526,
+      "learning_rate": 6.974768667831003e-06,
+      "loss": 0.4585,
+      "step": 1131
+    },
+    {
+      "epoch": 1.1163421247227014,
+      "grad_norm": 0.16484065818073312,
+      "learning_rate": 6.970015909201933e-06,
+      "loss": 0.46,
+      "step": 1132
+    },
+    {
+      "epoch": 1.1173280749322159,
+      "grad_norm": 0.16154224141757775,
+      "learning_rate": 6.965261042527967e-06,
+      "loss": 0.4618,
+      "step": 1133
+    },
+    {
+      "epoch": 1.1183140251417303,
+      "grad_norm": 0.17214462447951392,
+      "learning_rate": 6.960504072897119e-06,
+      "loss": 0.4623,
+      "step": 1134
+    },
+    {
+      "epoch": 1.1192999753512447,
+      "grad_norm": 0.1725997618736933,
+      "learning_rate": 6.9557450053996545e-06,
+      "loss": 0.4755,
+      "step": 1135
+    },
+    {
+      "epoch": 1.1202859255607591,
+      "grad_norm": 0.17276383842533038,
+      "learning_rate": 6.950983845128089e-06,
+      "loss": 0.4781,
+      "step": 1136
+    },
+    {
+      "epoch": 1.1212718757702735,
+      "grad_norm": 0.16154453911839164,
+      "learning_rate": 6.946220597177168e-06,
+      "loss": 0.4627,
+      "step": 1137
+    },
+    {
+      "epoch": 1.122257825979788,
+      "grad_norm": 0.1703763535245707,
+      "learning_rate": 6.94145526664388e-06,
+      "loss": 0.47,
+      "step": 1138
+    },
+    {
+      "epoch": 1.1232437761893024,
+      "grad_norm": 0.28774812472020417,
+      "learning_rate": 6.936687858627435e-06,
+      "loss": 0.4901,
+      "step": 1139
+    },
+    {
+      "epoch": 1.1242297263988168,
+      "grad_norm": 0.18344988614290553,
+      "learning_rate": 6.931918378229272e-06,
+      "loss": 0.479,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1252156766083312,
+      "grad_norm": 0.17717966653243214,
+      "learning_rate": 6.927146830553042e-06,
+      "loss": 0.4683,
+      "step": 1141
+    },
+    {
+      "epoch": 1.1262016268178456,
+      "grad_norm": 0.16444086161857938,
+      "learning_rate": 6.9223732207046135e-06,
+      "loss": 0.4628,
+      "step": 1142
+    },
+    {
+      "epoch": 1.12718757702736,
+      "grad_norm": 0.1820239289873006,
+      "learning_rate": 6.917597553792056e-06,
+      "loss": 0.4807,
+      "step": 1143
+    },
+    {
+      "epoch": 1.1281735272368745,
+      "grad_norm": 0.17719076702306055,
+      "learning_rate": 6.9128198349256425e-06,
+      "loss": 0.4749,
+      "step": 1144
+    },
+    {
+      "epoch": 1.129159477446389,
+      "grad_norm": 0.16908077087864143,
+      "learning_rate": 6.908040069217846e-06,
+      "loss": 0.4782,
+      "step": 1145
+    },
+    {
+      "epoch": 1.1301454276559033,
+      "grad_norm": 0.1746598197608833,
+      "learning_rate": 6.903258261783325e-06,
+      "loss": 0.4667,
+      "step": 1146
+    },
+    {
+      "epoch": 1.1311313778654177,
+      "grad_norm": 0.17432819355640805,
+      "learning_rate": 6.898474417738921e-06,
+      "loss": 0.4729,
+      "step": 1147
+    },
+    {
+      "epoch": 1.1321173280749321,
+      "grad_norm": 0.1793286884959327,
+      "learning_rate": 6.8936885422036605e-06,
+      "loss": 0.4644,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1331032782844466,
+      "grad_norm": 0.16360993038795005,
+      "learning_rate": 6.88890064029874e-06,
+      "loss": 0.4765,
+      "step": 1149
+    },
+    {
+      "epoch": 1.134089228493961,
+      "grad_norm": 0.21984283899142565,
+      "learning_rate": 6.884110717147524e-06,
+      "loss": 0.4668,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1350751787034754,
+      "grad_norm": 0.16589158210647748,
+      "learning_rate": 6.879318777875545e-06,
+      "loss": 0.4542,
+      "step": 1151
+    },
+    {
+      "epoch": 1.1360611289129898,
+      "grad_norm": 0.17202803684950171,
+      "learning_rate": 6.874524827610485e-06,
+      "loss": 0.4911,
+      "step": 1152
+    },
+    {
+      "epoch": 1.1370470791225042,
+      "grad_norm": 0.18297592552668762,
+      "learning_rate": 6.869728871482185e-06,
+      "loss": 0.4626,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1380330293320187,
+      "grad_norm": 0.1714123832013582,
+      "learning_rate": 6.864930914622627e-06,
+      "loss": 0.4708,
+      "step": 1154
+    },
+    {
+      "epoch": 1.139018979541533,
+      "grad_norm": 0.16642010529201842,
+      "learning_rate": 6.860130962165937e-06,
+      "loss": 0.4646,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1400049297510475,
+      "grad_norm": 0.16755274074404958,
+      "learning_rate": 6.855329019248377e-06,
+      "loss": 0.4543,
+      "step": 1156
+    },
+    {
+      "epoch": 1.140990879960562,
+      "grad_norm": 0.17173758815122947,
+      "learning_rate": 6.850525091008337e-06,
+      "loss": 0.4639,
+      "step": 1157
+    },
+    {
+      "epoch": 1.1419768301700763,
+      "grad_norm": 0.16844137773935142,
+      "learning_rate": 6.8457191825863305e-06,
+      "loss": 0.4876,
+      "step": 1158
+    },
+    {
+      "epoch": 1.1429627803795908,
+      "grad_norm": 0.17557928876543696,
+      "learning_rate": 6.840911299124993e-06,
+      "loss": 0.4696,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1439487305891052,
+      "grad_norm": 0.16525897262605324,
+      "learning_rate": 6.83610144576907e-06,
+      "loss": 0.4566,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1449346807986196,
+      "grad_norm": 0.16496392440325483,
+      "learning_rate": 6.831289627665418e-06,
+      "loss": 0.4689,
+      "step": 1161
+    },
+    {
+      "epoch": 1.145920631008134,
+      "grad_norm": 0.17009314190217278,
+      "learning_rate": 6.8264758499629966e-06,
+      "loss": 0.4886,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1469065812176484,
+      "grad_norm": 0.16660570181208462,
+      "learning_rate": 6.82166011781286e-06,
+      "loss": 0.4759,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1478925314271629,
+      "grad_norm": 0.16569799034550128,
+      "learning_rate": 6.816842436368152e-06,
+      "loss": 0.4705,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1488784816366773,
+      "grad_norm": 0.16786395945049576,
+      "learning_rate": 6.812022810784105e-06,
+      "loss": 0.4797,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1498644318461917,
+      "grad_norm": 0.17044498762909724,
+      "learning_rate": 6.807201246218032e-06,
+      "loss": 0.4849,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1508503820557061,
+      "grad_norm": 0.16634613228399914,
+      "learning_rate": 6.802377747829317e-06,
+      "loss": 0.4802,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1518363322652205,
+      "grad_norm": 0.17795643655428448,
+      "learning_rate": 6.7975523207794225e-06,
+      "loss": 0.4794,
+      "step": 1168
+    },
+    {
+      "epoch": 1.152822282474735,
+      "grad_norm": 0.1611271594226689,
+      "learning_rate": 6.792724970231863e-06,
+      "loss": 0.473,
+      "step": 1169
+    },
+    {
+      "epoch": 1.1538082326842494,
+      "grad_norm": 0.16386488268030752,
+      "learning_rate": 6.78789570135222e-06,
+      "loss": 0.4656,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1547941828937638,
+      "grad_norm": 0.15555865771083044,
+      "learning_rate": 6.783064519308124e-06,
+      "loss": 0.4513,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1557801331032782,
+      "grad_norm": 0.16149298292896075,
+      "learning_rate": 6.778231429269254e-06,
+      "loss": 0.4586,
+      "step": 1172
+    },
+    {
+      "epoch": 1.1567660833127926,
+      "grad_norm": 0.16270351048734,
+      "learning_rate": 6.773396436407329e-06,
+      "loss": 0.4597,
+      "step": 1173
+    },
+    {
+      "epoch": 1.157752033522307,
+      "grad_norm": 0.16464711130538048,
+      "learning_rate": 6.768559545896105e-06,
+      "loss": 0.4592,
+      "step": 1174
+    },
+    {
+      "epoch": 1.1587379837318215,
+      "grad_norm": 0.17311614033058068,
+      "learning_rate": 6.763720762911369e-06,
+      "loss": 0.4511,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1597239339413359,
+      "grad_norm": 0.16841372538790297,
+      "learning_rate": 6.758880092630935e-06,
+      "loss": 0.4743,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1607098841508503,
+      "grad_norm": 0.16587863597246358,
+      "learning_rate": 6.75403754023463e-06,
+      "loss": 0.4593,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1616958343603647,
+      "grad_norm": 0.1636018720957146,
+      "learning_rate": 6.749193110904303e-06,
+      "loss": 0.4797,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1626817845698791,
+      "grad_norm": 0.1740505271613513,
+      "learning_rate": 6.744346809823807e-06,
+      "loss": 0.4728,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1636677347793936,
+      "grad_norm": 0.2693536885187195,
+      "learning_rate": 6.739498642178999e-06,
+      "loss": 0.478,
+      "step": 1180
+    },
+    {
+      "epoch": 1.164653684988908,
+      "grad_norm": 0.16577868601241794,
+      "learning_rate": 6.734648613157732e-06,
+      "loss": 0.4624,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1656396351984224,
+      "grad_norm": 0.1744766961452551,
+      "learning_rate": 6.729796727949852e-06,
+      "loss": 0.4894,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1666255854079368,
+      "grad_norm": 0.16541529119470244,
+      "learning_rate": 6.724942991747191e-06,
+      "loss": 0.4584,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1676115356174512,
+      "grad_norm": 0.17414051379949466,
+      "learning_rate": 6.720087409743564e-06,
+      "loss": 0.4521,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1685974858269657,
+      "grad_norm": 0.1702068088669824,
+      "learning_rate": 6.715229987134757e-06,
+      "loss": 0.444,
+      "step": 1185
+    },
+    {
+      "epoch": 1.16958343603648,
+      "grad_norm": 0.16404748027845822,
+      "learning_rate": 6.710370729118527e-06,
+      "loss": 0.4677,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1705693862459945,
+      "grad_norm": 0.16056225532303414,
+      "learning_rate": 6.705509640894597e-06,
+      "loss": 0.46,
+      "step": 1187
+    },
+    {
+      "epoch": 1.171555336455509,
+      "grad_norm": 0.16207079229514923,
+      "learning_rate": 6.700646727664647e-06,
+      "loss": 0.467,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1725412866650233,
+      "grad_norm": 0.16149954151323434,
+      "learning_rate": 6.695781994632308e-06,
+      "loss": 0.4465,
+      "step": 1189
+    },
+    {
+      "epoch": 1.1735272368745377,
+      "grad_norm": 0.16566375036652037,
+      "learning_rate": 6.69091544700316e-06,
+      "loss": 0.4587,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1745131870840522,
+      "grad_norm": 0.16197493545476,
+      "learning_rate": 6.686047089984728e-06,
+      "loss": 0.4639,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1754991372935666,
+      "grad_norm": 0.17691894653661064,
+      "learning_rate": 6.681176928786467e-06,
+      "loss": 0.4581,
+      "step": 1192
+    },
+    {
+      "epoch": 1.176485087503081,
+      "grad_norm": 0.16291898757439355,
+      "learning_rate": 6.6763049686197665e-06,
+      "loss": 0.4419,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1774710377125954,
+      "grad_norm": 0.16579761031452891,
+      "learning_rate": 6.671431214697941e-06,
+      "loss": 0.4649,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1784569879221098,
+      "grad_norm": 0.16811772025638116,
+      "learning_rate": 6.666555672236222e-06,
+      "loss": 0.4809,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1794429381316243,
+      "grad_norm": 0.17605395928097767,
+      "learning_rate": 6.661678346451758e-06,
+      "loss": 0.4801,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1804288883411387,
+      "grad_norm": 0.16485846721219174,
+      "learning_rate": 6.656799242563603e-06,
+      "loss": 0.4724,
+      "step": 1197
+    },
+    {
+      "epoch": 1.181414838550653,
+      "grad_norm": 0.1631486631870248,
+      "learning_rate": 6.651918365792715e-06,
+      "loss": 0.4678,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1824007887601675,
+      "grad_norm": 0.1614771611712231,
+      "learning_rate": 6.647035721361951e-06,
+      "loss": 0.4919,
+      "step": 1199
+    },
+    {
+      "epoch": 1.183386738969682,
+      "grad_norm": 0.1652012155178517,
+      "learning_rate": 6.642151314496053e-06,
+      "loss": 0.4557,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1843726891791964,
+      "grad_norm": 0.16882925772854435,
+      "learning_rate": 6.637265150421658e-06,
+      "loss": 0.4986,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1853586393887108,
+      "grad_norm": 0.1628492998330556,
+      "learning_rate": 6.632377234367276e-06,
+      "loss": 0.4596,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1863445895982252,
+      "grad_norm": 0.26227874201540097,
+      "learning_rate": 6.627487571563293e-06,
+      "loss": 0.4763,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1873305398077396,
+      "grad_norm": 0.16499219832675732,
+      "learning_rate": 6.622596167241971e-06,
+      "loss": 0.4539,
+      "step": 1204
+    },
+    {
+      "epoch": 1.188316490017254,
+      "grad_norm": 0.2066692200887719,
+      "learning_rate": 6.617703026637426e-06,
+      "loss": 0.4734,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1893024402267685,
+      "grad_norm": 0.16630659374021256,
+      "learning_rate": 6.612808154985637e-06,
+      "loss": 0.4931,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1902883904362829,
+      "grad_norm": 0.17020363980871317,
+      "learning_rate": 6.607911557524434e-06,
+      "loss": 0.4779,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1912743406457973,
+      "grad_norm": 0.16697465970372505,
+      "learning_rate": 6.603013239493495e-06,
+      "loss": 0.4625,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1922602908553117,
+      "grad_norm": 0.17274524007780673,
+      "learning_rate": 6.598113206134338e-06,
+      "loss": 0.469,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1932462410648261,
+      "grad_norm": 0.15882550112893343,
+      "learning_rate": 6.593211462690317e-06,
+      "loss": 0.4672,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1942321912743405,
+      "grad_norm": 0.16581349881129095,
+      "learning_rate": 6.5883080144066145e-06,
+      "loss": 0.4793,
+      "step": 1211
+    },
+    {
+      "epoch": 1.195218141483855,
+      "grad_norm": 0.1713090589344652,
+      "learning_rate": 6.58340286653024e-06,
+      "loss": 0.4696,
+      "step": 1212
+    },
+    {
+      "epoch": 1.1962040916933696,
+      "grad_norm": 0.17539363763708982,
+      "learning_rate": 6.578496024310017e-06,
+      "loss": 0.4795,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1971900419028838,
+      "grad_norm": 0.16582363125647054,
+      "learning_rate": 6.573587492996589e-06,
+      "loss": 0.4806,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1981759921123984,
+      "grad_norm": 0.1749027122510544,
+      "learning_rate": 6.568677277842401e-06,
+      "loss": 0.459,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1991619423219126,
+      "grad_norm": 0.17937901370390394,
+      "learning_rate": 6.563765384101704e-06,
+      "loss": 0.475,
+      "step": 1216
+    },
+    {
+      "epoch": 1.2001478925314273,
+      "grad_norm": 0.17045662665449884,
+      "learning_rate": 6.558851817030541e-06,
+      "loss": 0.4719,
+      "step": 1217
+    },
+    {
+      "epoch": 1.2011338427409415,
+      "grad_norm": 0.1629843929250449,
+      "learning_rate": 6.5539365818867474e-06,
+      "loss": 0.4572,
+      "step": 1218
+    },
+    {
+      "epoch": 1.2021197929504561,
+      "grad_norm": 0.16442995828161608,
+      "learning_rate": 6.549019683929945e-06,
+      "loss": 0.4921,
+      "step": 1219
+    },
+    {
+      "epoch": 1.2031057431599703,
+      "grad_norm": 0.16487269651575245,
+      "learning_rate": 6.544101128421534e-06,
+      "loss": 0.4748,
+      "step": 1220
+    },
+    {
+      "epoch": 1.204091693369485,
+      "grad_norm": 0.16386476223301738,
+      "learning_rate": 6.539180920624687e-06,
+      "loss": 0.478,
+      "step": 1221
+    },
+    {
+      "epoch": 1.2050776435789992,
+      "grad_norm": 0.163091810991172,
+      "learning_rate": 6.534259065804348e-06,
+      "loss": 0.4411,
+      "step": 1222
+    },
+    {
+      "epoch": 1.2060635937885138,
+      "grad_norm": 0.16456812390268993,
+      "learning_rate": 6.5293355692272175e-06,
+      "loss": 0.4639,
+      "step": 1223
+    },
+    {
+      "epoch": 1.207049543998028,
+      "grad_norm": 0.1623734530086265,
+      "learning_rate": 6.52441043616176e-06,
+      "loss": 0.4813,
+      "step": 1224
+    },
+    {
+      "epoch": 1.2080354942075426,
+      "grad_norm": 0.19053264204778192,
+      "learning_rate": 6.519483671878184e-06,
+      "loss": 0.481,
+      "step": 1225
+    },
+    {
+      "epoch": 1.2090214444170568,
+      "grad_norm": 0.16596042008577788,
+      "learning_rate": 6.514555281648451e-06,
+      "loss": 0.4675,
+      "step": 1226
+    },
+    {
+      "epoch": 1.2100073946265715,
+      "grad_norm": 0.18419164498767882,
+      "learning_rate": 6.509625270746256e-06,
+      "loss": 0.4763,
+      "step": 1227
+    },
+    {
+      "epoch": 1.2109933448360857,
+      "grad_norm": 0.16272191942149414,
+      "learning_rate": 6.504693644447031e-06,
+      "loss": 0.4494,
+      "step": 1228
+    },
+    {
+      "epoch": 1.2119792950456003,
+      "grad_norm": 0.16270551892164925,
+      "learning_rate": 6.499760408027936e-06,
+      "loss": 0.4598,
+      "step": 1229
+    },
+    {
+      "epoch": 1.2129652452551145,
+      "grad_norm": 0.1847597471968673,
+      "learning_rate": 6.494825566767855e-06,
+      "loss": 0.4681,
+      "step": 1230
+    },
+    {
+      "epoch": 1.2139511954646292,
+      "grad_norm": 0.17468692693829316,
+      "learning_rate": 6.489889125947388e-06,
+      "loss": 0.4928,
+      "step": 1231
+    },
+    {
+      "epoch": 1.2149371456741433,
+      "grad_norm": 0.1713936637052054,
+      "learning_rate": 6.484951090848848e-06,
+      "loss": 0.4722,
+      "step": 1232
+    },
+    {
+      "epoch": 1.215923095883658,
+      "grad_norm": 0.1610173653813777,
+      "learning_rate": 6.480011466756251e-06,
+      "loss": 0.4705,
+      "step": 1233
+    },
+    {
+      "epoch": 1.2169090460931722,
+      "grad_norm": 0.16242661303668252,
+      "learning_rate": 6.475070258955317e-06,
+      "loss": 0.473,
+      "step": 1234
+    },
+    {
+      "epoch": 1.2178949963026868,
+      "grad_norm": 0.17018172840938867,
+      "learning_rate": 6.470127472733459e-06,
+      "loss": 0.4763,
+      "step": 1235
+    },
+    {
+      "epoch": 1.218880946512201,
+      "grad_norm": 0.16686134054815724,
+      "learning_rate": 6.465183113379778e-06,
+      "loss": 0.4659,
+      "step": 1236
+    },
+    {
+      "epoch": 1.2198668967217157,
+      "grad_norm": 0.16195163014916636,
+      "learning_rate": 6.4602371861850636e-06,
+      "loss": 0.461,
+      "step": 1237
+    },
+    {
+      "epoch": 1.2208528469312299,
+      "grad_norm": 0.1677888366893549,
+      "learning_rate": 6.455289696441772e-06,
+      "loss": 0.4503,
+      "step": 1238
+    },
+    {
+      "epoch": 1.2218387971407445,
+      "grad_norm": 0.16238862790861802,
+      "learning_rate": 6.450340649444045e-06,
+      "loss": 0.454,
+      "step": 1239
+    },
+    {
+      "epoch": 1.2228247473502587,
+      "grad_norm": 0.16927755272637426,
+      "learning_rate": 6.445390050487678e-06,
+      "loss": 0.4545,
+      "step": 1240
+    },
+    {
+      "epoch": 1.2238106975597733,
+      "grad_norm": 0.1772015943207728,
+      "learning_rate": 6.440437904870138e-06,
+      "loss": 0.477,
+      "step": 1241
+    },
+    {
+      "epoch": 1.2247966477692875,
+      "grad_norm": 0.163709112796267,
+      "learning_rate": 6.435484217890539e-06,
+      "loss": 0.4788,
+      "step": 1242
+    },
+    {
+      "epoch": 1.2257825979788022,
+      "grad_norm": 0.16481737010694317,
+      "learning_rate": 6.430528994849652e-06,
+      "loss": 0.4717,
+      "step": 1243
+    },
+    {
+      "epoch": 1.2267685481883164,
+      "grad_norm": 0.16548405260129903,
+      "learning_rate": 6.425572241049883e-06,
+      "loss": 0.4618,
+      "step": 1244
+    },
+    {
+      "epoch": 1.227754498397831,
+      "grad_norm": 0.16990397077693506,
+      "learning_rate": 6.420613961795284e-06,
+      "loss": 0.4742,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2287404486073452,
+      "grad_norm": 0.16030382598931203,
+      "learning_rate": 6.415654162391529e-06,
+      "loss": 0.4625,
+      "step": 1246
+    },
+    {
+      "epoch": 1.2297263988168599,
+      "grad_norm": 0.1674252777794334,
+      "learning_rate": 6.410692848145934e-06,
+      "loss": 0.4837,
+      "step": 1247
+    },
+    {
+      "epoch": 1.230712349026374,
+      "grad_norm": 0.16915053152864798,
+      "learning_rate": 6.40573002436742e-06,
+      "loss": 0.4801,
+      "step": 1248
+    },
+    {
+      "epoch": 1.2316982992358887,
+      "grad_norm": 0.16523840131205747,
+      "learning_rate": 6.4007656963665356e-06,
+      "loss": 0.4615,
+      "step": 1249
+    },
+    {
+      "epoch": 1.232684249445403,
+      "grad_norm": 0.15743869879284614,
+      "learning_rate": 6.395799869455433e-06,
+      "loss": 0.4577,
+      "step": 1250
+    },
+    {
+      "epoch": 1.2336701996549175,
+      "grad_norm": 0.15733935106047514,
+      "learning_rate": 6.390832548947866e-06,
+      "loss": 0.4599,
+      "step": 1251
+    },
+    {
+      "epoch": 1.2346561498644317,
+      "grad_norm": 0.1773091500925943,
+      "learning_rate": 6.385863740159194e-06,
+      "loss": 0.473,
+      "step": 1252
+    },
+    {
+      "epoch": 1.2356421000739464,
+      "grad_norm": 0.17114851438291118,
+      "learning_rate": 6.3808934484063625e-06,
+      "loss": 0.4749,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2366280502834606,
+      "grad_norm": 0.16869839701286612,
+      "learning_rate": 6.3759216790079085e-06,
+      "loss": 0.4648,
+      "step": 1254
+    },
+    {
+      "epoch": 1.2376140004929752,
+      "grad_norm": 0.16990904986951733,
+      "learning_rate": 6.370948437283944e-06,
+      "loss": 0.4629,
+      "step": 1255
+    },
+    {
+      "epoch": 1.2385999507024894,
+      "grad_norm": 0.17412698200049834,
+      "learning_rate": 6.365973728556164e-06,
+      "loss": 0.4851,
+      "step": 1256
+    },
+    {
+      "epoch": 1.239585900912004,
+      "grad_norm": 0.16591032990859195,
+      "learning_rate": 6.36099755814783e-06,
+      "loss": 0.4242,
+      "step": 1257
+    },
+    {
+      "epoch": 1.2405718511215182,
+      "grad_norm": 0.157269877576396,
+      "learning_rate": 6.3560199313837646e-06,
+      "loss": 0.4757,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2415578013310329,
+      "grad_norm": 0.18554747819079026,
+      "learning_rate": 6.351040853590354e-06,
+      "loss": 0.4481,
+      "step": 1259
+    },
+    {
+      "epoch": 1.242543751540547,
+      "grad_norm": 0.1648461295752464,
+      "learning_rate": 6.3460603300955334e-06,
+      "loss": 0.4565,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2435297017500617,
+      "grad_norm": 0.18926894216889145,
+      "learning_rate": 6.341078366228786e-06,
+      "loss": 0.4725,
+      "step": 1261
+    },
+    {
+      "epoch": 1.244515651959576,
+      "grad_norm": 0.16195265923026017,
+      "learning_rate": 6.336094967321138e-06,
+      "loss": 0.4685,
+      "step": 1262
+    },
+    {
+      "epoch": 1.2455016021690906,
+      "grad_norm": 0.1656917211614036,
+      "learning_rate": 6.331110138705148e-06,
+      "loss": 0.4779,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2464875523786048,
+      "grad_norm": 0.16142209518173187,
+      "learning_rate": 6.326123885714907e-06,
+      "loss": 0.5116,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2474735025881194,
+      "grad_norm": 0.31997566157907265,
+      "learning_rate": 6.32113621368603e-06,
+      "loss": 0.4564,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2484594527976336,
+      "grad_norm": 0.17730394323574275,
+      "learning_rate": 6.316147127955649e-06,
+      "loss": 0.4687,
+      "step": 1266
+    },
+    {
+      "epoch": 1.2494454030071482,
+      "grad_norm": 0.1620764753291925,
+      "learning_rate": 6.3111566338624095e-06,
+      "loss": 0.4594,
+      "step": 1267
+    },
+    {
+      "epoch": 1.2504313532166624,
+      "grad_norm": 0.1678496281527768,
+      "learning_rate": 6.306164736746464e-06,
+      "loss": 0.47,
+      "step": 1268
+    },
+    {
+      "epoch": 1.251417303426177,
+      "grad_norm": 0.16089754499247352,
+      "learning_rate": 6.3011714419494655e-06,
+      "loss": 0.4762,
+      "step": 1269
+    },
+    {
+      "epoch": 1.2524032536356913,
+      "grad_norm": 0.1672739021913524,
+      "learning_rate": 6.296176754814567e-06,
+      "loss": 0.4712,
+      "step": 1270
+    },
+    {
+      "epoch": 1.253389203845206,
+      "grad_norm": 0.16876494261501873,
+      "learning_rate": 6.291180680686404e-06,
+      "loss": 0.487,
+      "step": 1271
+    },
+    {
+      "epoch": 1.25437515405472,
+      "grad_norm": 0.16771770316083348,
+      "learning_rate": 6.2861832249111036e-06,
+      "loss": 0.4762,
+      "step": 1272
+    },
+    {
+      "epoch": 1.2553611042642348,
+      "grad_norm": 0.16924318609516403,
+      "learning_rate": 6.281184392836265e-06,
+      "loss": 0.4739,
+      "step": 1273
+    },
+    {
+      "epoch": 1.256347054473749,
+      "grad_norm": 0.1619535812048183,
+      "learning_rate": 6.276184189810964e-06,
+      "loss": 0.4793,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2573330046832636,
+      "grad_norm": 0.17080622465645712,
+      "learning_rate": 6.271182621185743e-06,
+      "loss": 0.4599,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2583189548927778,
+      "grad_norm": 0.16123342241345748,
+      "learning_rate": 6.266179692312604e-06,
+      "loss": 0.4461,
+      "step": 1276
+    },
+    {
+      "epoch": 1.2593049051022924,
+      "grad_norm": 0.16737822044918219,
+      "learning_rate": 6.261175408545007e-06,
+      "loss": 0.4578,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2602908553118066,
+      "grad_norm": 0.1703890577074199,
+      "learning_rate": 6.256169775237858e-06,
+      "loss": 0.4596,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2612768055213213,
+      "grad_norm": 0.16742321142966837,
+      "learning_rate": 6.251162797747513e-06,
+      "loss": 0.4679,
+      "step": 1279
+    },
+    {
+      "epoch": 1.2622627557308355,
+      "grad_norm": 0.1815832800104829,
+      "learning_rate": 6.246154481431761e-06,
+      "loss": 0.4718,
+      "step": 1280
+    },
+    {
+      "epoch": 1.26324870594035,
+      "grad_norm": 0.16788093578994845,
+      "learning_rate": 6.241144831649825e-06,
+      "loss": 0.4535,
+      "step": 1281
+    },
+    {
+      "epoch": 1.2642346561498643,
+      "grad_norm": 0.15990959895353038,
+      "learning_rate": 6.236133853762356e-06,
+      "loss": 0.4618,
+      "step": 1282
+    },
+    {
+      "epoch": 1.265220606359379,
+      "grad_norm": 0.16402331148916255,
+      "learning_rate": 6.2311215531314266e-06,
+      "loss": 0.4655,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2662065565688931,
+      "grad_norm": 0.1852919368904106,
+      "learning_rate": 6.226107935120521e-06,
+      "loss": 0.4718,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2671925067784078,
+      "grad_norm": 0.16327620865295686,
+      "learning_rate": 6.22109300509454e-06,
+      "loss": 0.4347,
+      "step": 1285
+    },
+    {
+      "epoch": 1.268178456987922,
+      "grad_norm": 0.15772623902003347,
+      "learning_rate": 6.216076768419782e-06,
+      "loss": 0.4558,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2691644071974366,
+      "grad_norm": 0.15816908510441025,
+      "learning_rate": 6.2110592304639465e-06,
+      "loss": 0.4716,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2701503574069508,
+      "grad_norm": 0.16483487579973904,
+      "learning_rate": 6.206040396596122e-06,
+      "loss": 0.4733,
+      "step": 1288
+    },
+    {
+      "epoch": 1.2711363076164655,
+      "grad_norm": 0.16045450842226178,
+      "learning_rate": 6.2010202721867905e-06,
+      "loss": 0.4708,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2721222578259797,
+      "grad_norm": 0.16640850424068243,
+      "learning_rate": 6.195998862607808e-06,
+      "loss": 0.4723,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2731082080354943,
+      "grad_norm": 0.16315540706301973,
+      "learning_rate": 6.190976173232411e-06,
+      "loss": 0.4691,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2740941582450085,
+      "grad_norm": 0.16461112623105786,
+      "learning_rate": 6.185952209435202e-06,
+      "loss": 0.448,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2750801084545231,
+      "grad_norm": 0.1612932926419046,
+      "learning_rate": 6.180926976592149e-06,
+      "loss": 0.4797,
+      "step": 1293
+    },
+    {
+      "epoch": 1.2760660586640373,
+      "grad_norm": 0.16633090978817794,
+      "learning_rate": 6.1759004800805745e-06,
+      "loss": 0.4737,
+      "step": 1294
+    },
+    {
+      "epoch": 1.277052008873552,
+      "grad_norm": 0.1668446080528459,
+      "learning_rate": 6.17087272527916e-06,
+      "loss": 0.4701,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2780379590830664,
+      "grad_norm": 0.16724989223783207,
+      "learning_rate": 6.165843717567928e-06,
+      "loss": 0.4626,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2790239092925808,
+      "grad_norm": 0.16331114452795373,
+      "learning_rate": 6.160813462328243e-06,
+      "loss": 0.4525,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2800098595020952,
+      "grad_norm": 0.1657302360887789,
+      "learning_rate": 6.155781964942805e-06,
+      "loss": 0.4708,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2809958097116096,
+      "grad_norm": 0.16514305117587627,
+      "learning_rate": 6.15074923079564e-06,
+      "loss": 0.474,
+      "step": 1299
+    },
+    {
+      "epoch": 1.281981759921124,
+      "grad_norm": 0.16147707778750775,
+      "learning_rate": 6.145715265272106e-06,
+      "loss": 0.4713,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2829677101306385,
+      "grad_norm": 0.16138878314330543,
+      "learning_rate": 6.140680073758868e-06,
+      "loss": 0.4553,
+      "step": 1301
+    },
+    {
+      "epoch": 1.283953660340153,
+      "grad_norm": 0.17315028127761717,
+      "learning_rate": 6.135643661643909e-06,
+      "loss": 0.4495,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2849396105496673,
+      "grad_norm": 0.1671419705848458,
+      "learning_rate": 6.1306060343165175e-06,
+      "loss": 0.4578,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2859255607591817,
+      "grad_norm": 0.16664147031192345,
+      "learning_rate": 6.125567197167281e-06,
+      "loss": 0.4685,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2869115109686962,
+      "grad_norm": 0.16788815671047838,
+      "learning_rate": 6.120527155588084e-06,
+      "loss": 0.4906,
+      "step": 1305
+    },
+    {
+      "epoch": 1.2878974611782106,
+      "grad_norm": 0.1666309231159149,
+      "learning_rate": 6.115485914972096e-06,
+      "loss": 0.4625,
+      "step": 1306
+    },
+    {
+      "epoch": 1.288883411387725,
+      "grad_norm": 0.16012722394551646,
+      "learning_rate": 6.110443480713771e-06,
+      "loss": 0.468,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2898693615972394,
+      "grad_norm": 0.15592975147241056,
+      "learning_rate": 6.1053998582088454e-06,
+      "loss": 0.4437,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2908553118067538,
+      "grad_norm": 0.16046644072262936,
+      "learning_rate": 6.1003550528543175e-06,
+      "loss": 0.4592,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2918412620162683,
+      "grad_norm": 0.1666441859359538,
+      "learning_rate": 6.0953090700484604e-06,
+      "loss": 0.4716,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2928272122257827,
+      "grad_norm": 0.16569836512630953,
+      "learning_rate": 6.0902619151908e-06,
+      "loss": 0.4803,
+      "step": 1311
+    },
+    {
+      "epoch": 1.293813162435297,
+      "grad_norm": 0.19211531217262628,
+      "learning_rate": 6.085213593682122e-06,
+      "loss": 0.4692,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2947991126448115,
+      "grad_norm": 0.163248231523702,
+      "learning_rate": 6.080164110924458e-06,
+      "loss": 0.4651,
+      "step": 1313
+    },
+    {
+      "epoch": 1.295785062854326,
+      "grad_norm": 0.17678443752305695,
+      "learning_rate": 6.07511347232108e-06,
+      "loss": 0.4803,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2967710130638403,
+      "grad_norm": 0.17048543008252076,
+      "learning_rate": 6.070061683276503e-06,
+      "loss": 0.472,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2977569632733548,
+      "grad_norm": 0.16168795263522812,
+      "learning_rate": 6.065008749196465e-06,
+      "loss": 0.4664,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2987429134828692,
+      "grad_norm": 0.16696176979258565,
+      "learning_rate": 6.0599546754879355e-06,
+      "loss": 0.4544,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2997288636923836,
+      "grad_norm": 0.1665789392319412,
+      "learning_rate": 6.054899467559101e-06,
+      "loss": 0.4628,
+      "step": 1318
+    },
+    {
+      "epoch": 1.300714813901898,
+      "grad_norm": 0.16431155738959483,
+      "learning_rate": 6.049843130819364e-06,
+      "loss": 0.4764,
+      "step": 1319
+    },
+    {
+      "epoch": 1.3017007641114124,
+      "grad_norm": 0.1631417297949849,
+      "learning_rate": 6.044785670679331e-06,
+      "loss": 0.4616,
+      "step": 1320
+    },
+    {
+      "epoch": 1.3026867143209269,
+      "grad_norm": 0.16076143236596774,
+      "learning_rate": 6.039727092550812e-06,
+      "loss": 0.4719,
+      "step": 1321
+    },
+    {
+      "epoch": 1.3036726645304413,
+      "grad_norm": 0.1604735184866197,
+      "learning_rate": 6.034667401846815e-06,
+      "loss": 0.456,
+      "step": 1322
+    },
+    {
+      "epoch": 1.3046586147399557,
+      "grad_norm": 0.1699803801824112,
+      "learning_rate": 6.02960660398154e-06,
+      "loss": 0.4766,
+      "step": 1323
+    },
+    {
+      "epoch": 1.3056445649494701,
+      "grad_norm": 0.1693299534355223,
+      "learning_rate": 6.024544704370364e-06,
+      "loss": 0.4537,
+      "step": 1324
+    },
+    {
+      "epoch": 1.3066305151589845,
+      "grad_norm": 0.16625676705170947,
+      "learning_rate": 6.019481708429853e-06,
+      "loss": 0.4947,
+      "step": 1325
+    },
+    {
+      "epoch": 1.307616465368499,
+      "grad_norm": 0.1897305463235135,
+      "learning_rate": 6.014417621577737e-06,
+      "loss": 0.4647,
+      "step": 1326
+    },
+    {
+      "epoch": 1.3086024155780134,
+      "grad_norm": 0.16692803122511632,
+      "learning_rate": 6.0093524492329216e-06,
+      "loss": 0.4721,
+      "step": 1327
+    },
+    {
+      "epoch": 1.3095883657875278,
+      "grad_norm": 0.20823083410362914,
+      "learning_rate": 6.004286196815467e-06,
+      "loss": 0.4639,
+      "step": 1328
+    },
+    {
+      "epoch": 1.3105743159970422,
+      "grad_norm": 0.1673197758356401,
+      "learning_rate": 5.999218869746595e-06,
+      "loss": 0.486,
+      "step": 1329
+    },
+    {
+      "epoch": 1.3115602662065566,
+      "grad_norm": 0.15998251611718828,
+      "learning_rate": 5.994150473448672e-06,
+      "loss": 0.4551,
+      "step": 1330
+    },
+    {
+      "epoch": 1.312546216416071,
+      "grad_norm": 0.1578637250927203,
+      "learning_rate": 5.989081013345211e-06,
+      "loss": 0.4536,
+      "step": 1331
+    },
+    {
+      "epoch": 1.3135321666255855,
+      "grad_norm": 0.16366294834189274,
+      "learning_rate": 5.984010494860865e-06,
+      "loss": 0.487,
+      "step": 1332
+    },
+    {
+      "epoch": 1.3145181168351,
+      "grad_norm": 0.1651012646941456,
+      "learning_rate": 5.978938923421418e-06,
+      "loss": 0.474,
+      "step": 1333
+    },
+    {
+      "epoch": 1.3155040670446143,
+      "grad_norm": 0.1721464729952915,
+      "learning_rate": 5.973866304453778e-06,
+      "loss": 0.4831,
+      "step": 1334
+    },
+    {
+      "epoch": 1.3164900172541287,
+      "grad_norm": 0.17333625543229284,
+      "learning_rate": 5.9687926433859785e-06,
+      "loss": 0.4833,
+      "step": 1335
+    },
+    {
+      "epoch": 1.3174759674636431,
+      "grad_norm": 0.16199175748787464,
+      "learning_rate": 5.963717945647167e-06,
+      "loss": 0.4589,
+      "step": 1336
+    },
+    {
+      "epoch": 1.3184619176731576,
+      "grad_norm": 0.1633274934497099,
+      "learning_rate": 5.958642216667598e-06,
+      "loss": 0.4635,
+      "step": 1337
+    },
+    {
+      "epoch": 1.319447867882672,
+      "grad_norm": 0.1647858407826564,
+      "learning_rate": 5.953565461878633e-06,
+      "loss": 0.4511,
+      "step": 1338
+    },
+    {
+      "epoch": 1.3204338180921864,
+      "grad_norm": 0.16666391657248938,
+      "learning_rate": 5.948487686712725e-06,
+      "loss": 0.4759,
+      "step": 1339
+    },
+    {
+      "epoch": 1.3214197683017008,
+      "grad_norm": 0.1608303742929351,
+      "learning_rate": 5.943408896603428e-06,
+      "loss": 0.4565,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3224057185112152,
+      "grad_norm": 0.16619810102543203,
+      "learning_rate": 5.938329096985374e-06,
+      "loss": 0.4877,
+      "step": 1341
+    },
+    {
+      "epoch": 1.3233916687207297,
+      "grad_norm": 0.16861142230966175,
+      "learning_rate": 5.933248293294278e-06,
+      "loss": 0.4817,
+      "step": 1342
+    },
+    {
+      "epoch": 1.324377618930244,
+      "grad_norm": 0.1613311991031562,
+      "learning_rate": 5.928166490966933e-06,
+      "loss": 0.4658,
+      "step": 1343
+    },
+    {
+      "epoch": 1.3253635691397585,
+      "grad_norm": 0.16613527042960732,
+      "learning_rate": 5.923083695441193e-06,
+      "loss": 0.4797,
+      "step": 1344
+    },
+    {
+      "epoch": 1.326349519349273,
+      "grad_norm": 0.17375547190336268,
+      "learning_rate": 5.9179999121559816e-06,
+      "loss": 0.4644,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3273354695587873,
+      "grad_norm": 0.16483341515761354,
+      "learning_rate": 5.912915146551278e-06,
+      "loss": 0.4826,
+      "step": 1346
+    },
+    {
+      "epoch": 1.3283214197683018,
+      "grad_norm": 0.1645542396832144,
+      "learning_rate": 5.907829404068108e-06,
+      "loss": 0.468,
+      "step": 1347
+    },
+    {
+      "epoch": 1.3293073699778162,
+      "grad_norm": 0.1671629788898561,
+      "learning_rate": 5.902742690148551e-06,
+      "loss": 0.486,
+      "step": 1348
+    },
+    {
+      "epoch": 1.3302933201873306,
+      "grad_norm": 0.16099037162037994,
+      "learning_rate": 5.897655010235715e-06,
+      "loss": 0.4563,
+      "step": 1349
+    },
+    {
+      "epoch": 1.331279270396845,
+      "grad_norm": 0.18211649634751956,
+      "learning_rate": 5.892566369773753e-06,
+      "loss": 0.4551,
+      "step": 1350
+    },
+    {
+      "epoch": 1.3322652206063594,
+      "grad_norm": 0.16220967017743546,
+      "learning_rate": 5.887476774207839e-06,
+      "loss": 0.4642,
+      "step": 1351
+    },
+    {
+      "epoch": 1.3332511708158739,
+      "grad_norm": 0.19336714192891388,
+      "learning_rate": 5.88238622898417e-06,
+      "loss": 0.457,
+      "step": 1352
+    },
+    {
+      "epoch": 1.3342371210253883,
+      "grad_norm": 0.1625183100395215,
+      "learning_rate": 5.87729473954996e-06,
+      "loss": 0.4542,
+      "step": 1353
+    },
+    {
+      "epoch": 1.3352230712349027,
+      "grad_norm": 0.16533821732733284,
+      "learning_rate": 5.872202311353433e-06,
+      "loss": 0.4669,
+      "step": 1354
+    },
+    {
+      "epoch": 1.336209021444417,
+      "grad_norm": 0.16833781950905896,
+      "learning_rate": 5.867108949843817e-06,
+      "loss": 0.4893,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3371949716539315,
+      "grad_norm": 0.16855094056463374,
+      "learning_rate": 5.8620146604713435e-06,
+      "loss": 0.4774,
+      "step": 1356
+    },
+    {
+      "epoch": 1.338180921863446,
+      "grad_norm": 0.16556996608166277,
+      "learning_rate": 5.856919448687226e-06,
+      "loss": 0.4715,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3391668720729604,
+      "grad_norm": 0.16669043040756595,
+      "learning_rate": 5.851823319943678e-06,
+      "loss": 0.4788,
+      "step": 1358
+    },
+    {
+      "epoch": 1.3401528222824748,
+      "grad_norm": 0.16574750712062733,
+      "learning_rate": 5.846726279693885e-06,
+      "loss": 0.4704,
+      "step": 1359
+    },
+    {
+      "epoch": 1.3411387724919892,
+      "grad_norm": 0.17279517228358982,
+      "learning_rate": 5.841628333392011e-06,
+      "loss": 0.472,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3421247227015036,
+      "grad_norm": 0.1630485863540864,
+      "learning_rate": 5.836529486493191e-06,
+      "loss": 0.4632,
+      "step": 1361
+    },
+    {
+      "epoch": 1.343110672911018,
+      "grad_norm": 0.1633776581966279,
+      "learning_rate": 5.831429744453519e-06,
+      "loss": 0.4771,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3440966231205325,
+      "grad_norm": 0.1666344556896262,
+      "learning_rate": 5.826329112730056e-06,
+      "loss": 0.4446,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3450825733300469,
+      "grad_norm": 0.16326008063661362,
+      "learning_rate": 5.821227596780802e-06,
+      "loss": 0.4713,
+      "step": 1364
+    },
+    {
+      "epoch": 1.3460685235395613,
+      "grad_norm": 0.16760883043753336,
+      "learning_rate": 5.816125202064714e-06,
+      "loss": 0.4734,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3470544737490757,
+      "grad_norm": 0.16263318856999043,
+      "learning_rate": 5.811021934041685e-06,
+      "loss": 0.4733,
+      "step": 1366
+    },
+    {
+      "epoch": 1.3480404239585901,
+      "grad_norm": 0.16694734396059974,
+      "learning_rate": 5.805917798172543e-06,
+      "loss": 0.4625,
+      "step": 1367
+    },
+    {
+      "epoch": 1.3490263741681046,
+      "grad_norm": 0.15700598417509207,
+      "learning_rate": 5.800812799919046e-06,
+      "loss": 0.4603,
+      "step": 1368
+    },
+    {
+      "epoch": 1.350012324377619,
+      "grad_norm": 0.1640378078382763,
+      "learning_rate": 5.795706944743871e-06,
+      "loss": 0.4681,
+      "step": 1369
+    },
+    {
+      "epoch": 1.3509982745871334,
+      "grad_norm": 0.16548375785466893,
+      "learning_rate": 5.790600238110614e-06,
+      "loss": 0.4625,
+      "step": 1370
+    },
+    {
+      "epoch": 1.3519842247966478,
+      "grad_norm": 0.16307590431256164,
+      "learning_rate": 5.785492685483787e-06,
+      "loss": 0.4484,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3529701750061622,
+      "grad_norm": 0.16742781126574774,
+      "learning_rate": 5.780384292328798e-06,
+      "loss": 0.4803,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3539561252156767,
+      "grad_norm": 0.1613501885215972,
+      "learning_rate": 5.775275064111962e-06,
+      "loss": 0.4604,
+      "step": 1373
+    },
+    {
+      "epoch": 1.354942075425191,
+      "grad_norm": 0.19920118243153734,
+      "learning_rate": 5.770165006300485e-06,
+      "loss": 0.4681,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3559280256347055,
+      "grad_norm": 0.1594088819002285,
+      "learning_rate": 5.765054124362458e-06,
+      "loss": 0.4678,
+      "step": 1375
+    },
+    {
+      "epoch": 1.35691397584422,
+      "grad_norm": 0.1700686727742075,
+      "learning_rate": 5.759942423766859e-06,
+      "loss": 0.4842,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3578999260537343,
+      "grad_norm": 0.1655330450181918,
+      "learning_rate": 5.754829909983539e-06,
+      "loss": 0.4704,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3588858762632487,
+      "grad_norm": 0.16107473023925153,
+      "learning_rate": 5.7497165884832185e-06,
+      "loss": 0.4619,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3598718264727632,
+      "grad_norm": 0.16500048672225306,
+      "learning_rate": 5.744602464737484e-06,
+      "loss": 0.4755,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3608577766822776,
+      "grad_norm": 0.20843416692703762,
+      "learning_rate": 5.739487544218779e-06,
+      "loss": 0.4523,
+      "step": 1380
+    },
+    {
+      "epoch": 1.361843726891792,
+      "grad_norm": 0.16443032992627848,
+      "learning_rate": 5.734371832400403e-06,
+      "loss": 0.4784,
+      "step": 1381
+    },
+    {
+      "epoch": 1.3628296771013064,
+      "grad_norm": 0.17828470247210604,
+      "learning_rate": 5.729255334756497e-06,
+      "loss": 0.4719,
+      "step": 1382
+    },
+    {
+      "epoch": 1.3638156273108208,
+      "grad_norm": 0.1649213056572908,
+      "learning_rate": 5.7241380567620475e-06,
+      "loss": 0.4643,
+      "step": 1383
+    },
+    {
+      "epoch": 1.3648015775203353,
+      "grad_norm": 0.1633401231658992,
+      "learning_rate": 5.719020003892873e-06,
+      "loss": 0.4676,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3657875277298497,
+      "grad_norm": 0.16414710207524366,
+      "learning_rate": 5.7139011816256215e-06,
+      "loss": 0.4531,
+      "step": 1385
+    },
+    {
+      "epoch": 1.366773477939364,
+      "grad_norm": 0.1861612278327007,
+      "learning_rate": 5.708781595437769e-06,
+      "loss": 0.4669,
+      "step": 1386
+    },
+    {
+      "epoch": 1.3677594281488785,
+      "grad_norm": 0.16442997293411388,
+      "learning_rate": 5.703661250807599e-06,
+      "loss": 0.4622,
+      "step": 1387
+    },
+    {
+      "epoch": 1.368745378358393,
+      "grad_norm": 0.16661472250606849,
+      "learning_rate": 5.698540153214218e-06,
+      "loss": 0.4687,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3697313285679074,
+      "grad_norm": 0.16345551302490932,
+      "learning_rate": 5.69341830813753e-06,
+      "loss": 0.475,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3707172787774218,
+      "grad_norm": 0.15839635099473848,
+      "learning_rate": 5.688295721058242e-06,
+      "loss": 0.4632,
+      "step": 1390
+    },
+    {
+      "epoch": 1.3717032289869362,
+      "grad_norm": 0.1622382829800274,
+      "learning_rate": 5.683172397457856e-06,
+      "loss": 0.4537,
+      "step": 1391
+    },
+    {
+      "epoch": 1.3726891791964506,
+      "grad_norm": 0.18459014148936892,
+      "learning_rate": 5.678048342818658e-06,
+      "loss": 0.4481,
+      "step": 1392
+    },
+    {
+      "epoch": 1.373675129405965,
+      "grad_norm": 0.1697886021302356,
+      "learning_rate": 5.672923562623722e-06,
+      "loss": 0.4617,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3746610796154795,
+      "grad_norm": 0.1645550056439435,
+      "learning_rate": 5.667798062356895e-06,
+      "loss": 0.4577,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3756470298249939,
+      "grad_norm": 0.17143846073074367,
+      "learning_rate": 5.662671847502793e-06,
+      "loss": 0.4523,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3766329800345083,
+      "grad_norm": 0.16344128759827684,
+      "learning_rate": 5.657544923546803e-06,
+      "loss": 0.4666,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3776189302440227,
+      "grad_norm": 0.16425518034491454,
+      "learning_rate": 5.65241729597506e-06,
+      "loss": 0.4638,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3786048804535371,
+      "grad_norm": 0.16534170302161283,
+      "learning_rate": 5.647288970274463e-06,
+      "loss": 0.4511,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3795908306630515,
+      "grad_norm": 0.1617323018590922,
+      "learning_rate": 5.642159951932652e-06,
+      "loss": 0.4702,
+      "step": 1399
+    },
+    {
+      "epoch": 1.380576780872566,
+      "grad_norm": 0.16083214099638724,
+      "learning_rate": 5.63703024643801e-06,
+      "loss": 0.4828,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3815627310820804,
+      "grad_norm": 0.1720696143022807,
+      "learning_rate": 5.631899859279654e-06,
+      "loss": 0.475,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3825486812915948,
+      "grad_norm": 0.17435364748236737,
+      "learning_rate": 5.626768795947432e-06,
+      "loss": 0.4802,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3835346315011092,
+      "grad_norm": 0.16437688195121364,
+      "learning_rate": 5.6216370619319134e-06,
+      "loss": 0.4609,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3845205817106236,
+      "grad_norm": 0.16548132070839128,
+      "learning_rate": 5.61650466272439e-06,
+      "loss": 0.484,
+      "step": 1404
+    },
+    {
+      "epoch": 1.385506531920138,
+      "grad_norm": 0.1549013795519419,
+      "learning_rate": 5.61137160381686e-06,
+      "loss": 0.4506,
+      "step": 1405
+    },
+    {
+      "epoch": 1.3864924821296525,
+      "grad_norm": 0.16939320130209762,
+      "learning_rate": 5.606237890702028e-06,
+      "loss": 0.4718,
+      "step": 1406
+    },
+    {
+      "epoch": 1.387478432339167,
+      "grad_norm": 0.17051233592675338,
+      "learning_rate": 5.601103528873304e-06,
+      "loss": 0.4367,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3884643825486813,
+      "grad_norm": 0.15943541732452363,
+      "learning_rate": 5.595968523824784e-06,
+      "loss": 0.4704,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3894503327581957,
+      "grad_norm": 0.16329645709213123,
+      "learning_rate": 5.590832881051262e-06,
+      "loss": 0.4768,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3904362829677102,
+      "grad_norm": 0.15557702222562347,
+      "learning_rate": 5.5856966060482024e-06,
+      "loss": 0.4647,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3914222331772246,
+      "grad_norm": 0.1561786248276199,
+      "learning_rate": 5.58055970431176e-06,
+      "loss": 0.4519,
+      "step": 1411
+    },
+    {
+      "epoch": 1.392408183386739,
+      "grad_norm": 0.16384753935726618,
+      "learning_rate": 5.575422181338748e-06,
+      "loss": 0.4632,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3933941335962534,
+      "grad_norm": 0.15968604821114685,
+      "learning_rate": 5.570284042626651e-06,
+      "loss": 0.4497,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3943800838057678,
+      "grad_norm": 0.16620749209474933,
+      "learning_rate": 5.565145293673612e-06,
+      "loss": 0.4656,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3953660340152823,
+      "grad_norm": 0.1669612046116747,
+      "learning_rate": 5.5600059399784245e-06,
+      "loss": 0.4662,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3963519842247967,
+      "grad_norm": 0.17082902317335172,
+      "learning_rate": 5.554865987040532e-06,
+      "loss": 0.468,
+      "step": 1416
+    },
+    {
+      "epoch": 1.397337934434311,
+      "grad_norm": 0.15940364329031975,
+      "learning_rate": 5.549725440360016e-06,
+      "loss": 0.4485,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3983238846438255,
+      "grad_norm": 0.16212216172085825,
+      "learning_rate": 5.5445843054375945e-06,
+      "loss": 0.4583,
+      "step": 1418
+    },
+    {
+      "epoch": 1.39930983485334,
+      "grad_norm": 0.163355919009208,
+      "learning_rate": 5.53944258777462e-06,
+      "loss": 0.4568,
+      "step": 1419
+    },
+    {
+      "epoch": 1.4002957850628543,
+      "grad_norm": 0.16262161805177902,
+      "learning_rate": 5.534300292873059e-06,
+      "loss": 0.4512,
+      "step": 1420
+    },
+    {
+      "epoch": 1.4012817352723688,
+      "grad_norm": 0.16106454629537711,
+      "learning_rate": 5.5291574262355055e-06,
+      "loss": 0.4613,
+      "step": 1421
+    },
+    {
+      "epoch": 1.4022676854818832,
+      "grad_norm": 0.184319064463163,
+      "learning_rate": 5.524013993365156e-06,
+      "loss": 0.4565,
+      "step": 1422
+    },
+    {
+      "epoch": 1.4032536356913976,
+      "grad_norm": 0.17067281877077936,
+      "learning_rate": 5.518869999765821e-06,
+      "loss": 0.4523,
+      "step": 1423
+    },
+    {
+      "epoch": 1.404239585900912,
+      "grad_norm": 0.1653886506783148,
+      "learning_rate": 5.513725450941906e-06,
+      "loss": 0.4795,
+      "step": 1424
+    },
+    {
+      "epoch": 1.4052255361104264,
+      "grad_norm": 0.1738770503738012,
+      "learning_rate": 5.508580352398413e-06,
+      "loss": 0.4845,
+      "step": 1425
+    },
+    {
+      "epoch": 1.4062114863199409,
+      "grad_norm": 0.16058762206241256,
+      "learning_rate": 5.503434709640929e-06,
+      "loss": 0.4586,
+      "step": 1426
+    },
+    {
+      "epoch": 1.4071974365294553,
+      "grad_norm": 0.16389304571758812,
+      "learning_rate": 5.498288528175628e-06,
+      "loss": 0.4505,
+      "step": 1427
+    },
+    {
+      "epoch": 1.4081833867389697,
+      "grad_norm": 0.16483781266044475,
+      "learning_rate": 5.49314181350926e-06,
+      "loss": 0.4583,
+      "step": 1428
+    },
+    {
+      "epoch": 1.4091693369484841,
+      "grad_norm": 0.16977386868298516,
+      "learning_rate": 5.487994571149139e-06,
+      "loss": 0.4735,
+      "step": 1429
+    },
+    {
+      "epoch": 1.4101552871579985,
+      "grad_norm": 0.17373734902192936,
+      "learning_rate": 5.482846806603153e-06,
+      "loss": 0.4739,
+      "step": 1430
+    },
+    {
+      "epoch": 1.411141237367513,
+      "grad_norm": 0.15581196161677877,
+      "learning_rate": 5.47769852537974e-06,
+      "loss": 0.4747,
+      "step": 1431
+    },
+    {
+      "epoch": 1.4121271875770274,
+      "grad_norm": 0.18696441199681657,
+      "learning_rate": 5.4725497329879006e-06,
+      "loss": 0.4656,
+      "step": 1432
+    },
+    {
+      "epoch": 1.4131131377865418,
+      "grad_norm": 0.16242538336812662,
+      "learning_rate": 5.46740043493717e-06,
+      "loss": 0.4763,
+      "step": 1433
+    },
+    {
+      "epoch": 1.4140990879960562,
+      "grad_norm": 0.16208367941537546,
+      "learning_rate": 5.462250636737638e-06,
+      "loss": 0.4671,
+      "step": 1434
+    },
+    {
+      "epoch": 1.4150850382055706,
+      "grad_norm": 0.1681451829161332,
+      "learning_rate": 5.457100343899918e-06,
+      "loss": 0.4628,
+      "step": 1435
+    },
+    {
+      "epoch": 1.416070988415085,
+      "grad_norm": 0.1606132001352093,
+      "learning_rate": 5.451949561935161e-06,
+      "loss": 0.4793,
+      "step": 1436
+    },
+    {
+      "epoch": 1.4170569386245995,
+      "grad_norm": 0.1601840747483185,
+      "learning_rate": 5.4467982963550346e-06,
+      "loss": 0.4566,
+      "step": 1437
+    },
+    {
+      "epoch": 1.418042888834114,
+      "grad_norm": 0.16071872559154282,
+      "learning_rate": 5.441646552671731e-06,
+      "loss": 0.4583,
+      "step": 1438
+    },
+    {
+      "epoch": 1.4190288390436283,
+      "grad_norm": 0.16357197680596194,
+      "learning_rate": 5.436494336397948e-06,
+      "loss": 0.4604,
+      "step": 1439
+    },
+    {
+      "epoch": 1.4200147892531427,
+      "grad_norm": 0.168978827594285,
+      "learning_rate": 5.431341653046893e-06,
+      "loss": 0.4775,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4210007394626571,
+      "grad_norm": 0.16579442234725897,
+      "learning_rate": 5.4261885081322685e-06,
+      "loss": 0.4535,
+      "step": 1441
+    },
+    {
+      "epoch": 1.4219866896721716,
+      "grad_norm": 0.1577143915266565,
+      "learning_rate": 5.421034907168279e-06,
+      "loss": 0.4651,
+      "step": 1442
+    },
+    {
+      "epoch": 1.422972639881686,
+      "grad_norm": 0.16462619439183032,
+      "learning_rate": 5.415880855669607e-06,
+      "loss": 0.4716,
+      "step": 1443
+    },
+    {
+      "epoch": 1.4239585900912004,
+      "grad_norm": 0.15661772083930173,
+      "learning_rate": 5.410726359151426e-06,
+      "loss": 0.4581,
+      "step": 1444
+    },
+    {
+      "epoch": 1.4249445403007148,
+      "grad_norm": 0.15937762868230831,
+      "learning_rate": 5.40557142312938e-06,
+      "loss": 0.4811,
+      "step": 1445
+    },
+    {
+      "epoch": 1.4259304905102292,
+      "grad_norm": 0.16770318868465095,
+      "learning_rate": 5.400416053119586e-06,
+      "loss": 0.4624,
+      "step": 1446
+    },
+    {
+      "epoch": 1.4269164407197437,
+      "grad_norm": 0.15556893949406567,
+      "learning_rate": 5.395260254638624e-06,
+      "loss": 0.4672,
+      "step": 1447
+    },
+    {
+      "epoch": 1.427902390929258,
+      "grad_norm": 0.16142586376021115,
+      "learning_rate": 5.390104033203533e-06,
+      "loss": 0.4537,
+      "step": 1448
+    },
+    {
+      "epoch": 1.4288883411387725,
+      "grad_norm": 0.1650474674463431,
+      "learning_rate": 5.3849473943318045e-06,
+      "loss": 0.473,
+      "step": 1449
+    },
+    {
+      "epoch": 1.429874291348287,
+      "grad_norm": 0.16021595164370292,
+      "learning_rate": 5.379790343541376e-06,
+      "loss": 0.456,
+      "step": 1450
+    },
+    {
+      "epoch": 1.4308602415578013,
+      "grad_norm": 0.16147362832111686,
+      "learning_rate": 5.374632886350628e-06,
+      "loss": 0.4627,
+      "step": 1451
+    },
+    {
+      "epoch": 1.4318461917673158,
+      "grad_norm": 0.16205571482787692,
+      "learning_rate": 5.3694750282783745e-06,
+      "loss": 0.4455,
+      "step": 1452
+    },
+    {
+      "epoch": 1.4328321419768302,
+      "grad_norm": 0.15866834834604854,
+      "learning_rate": 5.36431677484386e-06,
+      "loss": 0.466,
+      "step": 1453
+    },
+    {
+      "epoch": 1.4338180921863446,
+      "grad_norm": 0.17128940295948444,
+      "learning_rate": 5.3591581315667465e-06,
+      "loss": 0.4717,
+      "step": 1454
+    },
+    {
+      "epoch": 1.434804042395859,
+      "grad_norm": 0.1661069463954318,
+      "learning_rate": 5.353999103967119e-06,
+      "loss": 0.4846,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4357899926053734,
+      "grad_norm": 0.16101274669849455,
+      "learning_rate": 5.348839697565472e-06,
+      "loss": 0.4547,
+      "step": 1456
+    },
+    {
+      "epoch": 1.4367759428148879,
+      "grad_norm": 0.1576415609417486,
+      "learning_rate": 5.343679917882707e-06,
+      "loss": 0.4614,
+      "step": 1457
+    },
+    {
+      "epoch": 1.4377618930244023,
+      "grad_norm": 0.17253401627812268,
+      "learning_rate": 5.338519770440119e-06,
+      "loss": 0.4699,
+      "step": 1458
+    },
+    {
+      "epoch": 1.4387478432339167,
+      "grad_norm": 0.16345541147492304,
+      "learning_rate": 5.333359260759406e-06,
+      "loss": 0.4583,
+      "step": 1459
+    },
+    {
+      "epoch": 1.439733793443431,
+      "grad_norm": 0.17320027594932633,
+      "learning_rate": 5.3281983943626436e-06,
+      "loss": 0.4754,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4407197436529455,
+      "grad_norm": 0.16554097410343635,
+      "learning_rate": 5.3230371767722966e-06,
+      "loss": 0.4756,
+      "step": 1461
+    },
+    {
+      "epoch": 1.44170569386246,
+      "grad_norm": 0.1660296139253079,
+      "learning_rate": 5.317875613511202e-06,
+      "loss": 0.4574,
+      "step": 1462
+    },
+    {
+      "epoch": 1.4426916440719744,
+      "grad_norm": 0.1617940600058465,
+      "learning_rate": 5.312713710102567e-06,
+      "loss": 0.4615,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4436775942814888,
+      "grad_norm": 0.16950274075793623,
+      "learning_rate": 5.307551472069964e-06,
+      "loss": 0.4792,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4446635444910032,
+      "grad_norm": 0.17120408352777172,
+      "learning_rate": 5.302388904937323e-06,
+      "loss": 0.4708,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4456494947005176,
+      "grad_norm": 0.1645352322140342,
+      "learning_rate": 5.2972260142289255e-06,
+      "loss": 0.4603,
+      "step": 1466
+    },
+    {
+      "epoch": 1.446635444910032,
+      "grad_norm": 0.16716175595707267,
+      "learning_rate": 5.2920628054694004e-06,
+      "loss": 0.4691,
+      "step": 1467
+    },
+    {
+      "epoch": 1.4476213951195465,
+      "grad_norm": 0.165255412064843,
+      "learning_rate": 5.286899284183714e-06,
+      "loss": 0.4617,
+      "step": 1468
+    },
+    {
+      "epoch": 1.4486073453290609,
+      "grad_norm": 0.1605534962378199,
+      "learning_rate": 5.281735455897172e-06,
+      "loss": 0.457,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4495932955385753,
+      "grad_norm": 0.1620741209304123,
+      "learning_rate": 5.276571326135405e-06,
+      "loss": 0.4545,
+      "step": 1470
+    },
+    {
+      "epoch": 1.4505792457480897,
+      "grad_norm": 0.15537470368008646,
+      "learning_rate": 5.271406900424366e-06,
+      "loss": 0.4581,
+      "step": 1471
+    },
+    {
+      "epoch": 1.4515651959576041,
+      "grad_norm": 0.1606369928134782,
+      "learning_rate": 5.266242184290327e-06,
+      "loss": 0.4704,
+      "step": 1472
+    },
+    {
+      "epoch": 1.4525511461671186,
+      "grad_norm": 0.16149958580223164,
+      "learning_rate": 5.261077183259867e-06,
+      "loss": 0.4487,
+      "step": 1473
+    },
+    {
+      "epoch": 1.453537096376633,
+      "grad_norm": 0.1639900897443875,
+      "learning_rate": 5.2559119028598775e-06,
+      "loss": 0.4466,
+      "step": 1474
+    },
+    {
+      "epoch": 1.4545230465861474,
+      "grad_norm": 0.16965276827089013,
+      "learning_rate": 5.250746348617538e-06,
+      "loss": 0.4743,
+      "step": 1475
+    },
+    {
+      "epoch": 1.4555089967956618,
+      "grad_norm": 0.16501635832606454,
+      "learning_rate": 5.245580526060331e-06,
+      "loss": 0.446,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4564949470051762,
+      "grad_norm": 0.15996363943632896,
+      "learning_rate": 5.2404144407160195e-06,
+      "loss": 0.4678,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4574808972146907,
+      "grad_norm": 0.16159558576747365,
+      "learning_rate": 5.235248098112652e-06,
+      "loss": 0.4726,
+      "step": 1478
+    },
+    {
+      "epoch": 1.458466847424205,
+      "grad_norm": 0.16461503170307748,
+      "learning_rate": 5.230081503778548e-06,
+      "loss": 0.4693,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4594527976337195,
+      "grad_norm": 0.15563785455833773,
+      "learning_rate": 5.224914663242303e-06,
+      "loss": 0.4559,
+      "step": 1480
+    },
+    {
+      "epoch": 1.460438747843234,
+      "grad_norm": 0.1651776006276481,
+      "learning_rate": 5.219747582032767e-06,
+      "loss": 0.4746,
+      "step": 1481
+    },
+    {
+      "epoch": 1.4614246980527483,
+      "grad_norm": 0.16077563836492886,
+      "learning_rate": 5.214580265679055e-06,
+      "loss": 0.46,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4624106482622627,
+      "grad_norm": 0.16175140377905586,
+      "learning_rate": 5.209412719710529e-06,
+      "loss": 0.4657,
+      "step": 1483
+    },
+    {
+      "epoch": 1.4633965984717772,
+      "grad_norm": 0.1696886394681388,
+      "learning_rate": 5.204244949656802e-06,
+      "loss": 0.4611,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4643825486812916,
+      "grad_norm": 0.15941472334718274,
+      "learning_rate": 5.19907696104772e-06,
+      "loss": 0.487,
+      "step": 1485
+    },
+    {
+      "epoch": 1.465368498890806,
+      "grad_norm": 0.16204398737122389,
+      "learning_rate": 5.193908759413369e-06,
+      "loss": 0.4795,
+      "step": 1486
+    },
+    {
+      "epoch": 1.4663544491003204,
+      "grad_norm": 0.1658665941820137,
+      "learning_rate": 5.188740350284058e-06,
+      "loss": 0.4766,
+      "step": 1487
+    },
+    {
+      "epoch": 1.4673403993098348,
+      "grad_norm": 0.16443166686761135,
+      "learning_rate": 5.18357173919032e-06,
+      "loss": 0.4696,
+      "step": 1488
+    },
+    {
+      "epoch": 1.4683263495193493,
+      "grad_norm": 0.1961560666508905,
+      "learning_rate": 5.178402931662905e-06,
+      "loss": 0.4579,
+      "step": 1489
+    },
+    {
+      "epoch": 1.4693122997288637,
+      "grad_norm": 0.16138047883085868,
+      "learning_rate": 5.173233933232774e-06,
+      "loss": 0.4583,
+      "step": 1490
+    },
+    {
+      "epoch": 1.470298249938378,
+      "grad_norm": 0.16087565868089423,
+      "learning_rate": 5.168064749431089e-06,
+      "loss": 0.4507,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4712842001478925,
+      "grad_norm": 0.15835325164556538,
+      "learning_rate": 5.162895385789214e-06,
+      "loss": 0.4618,
+      "step": 1492
+    },
+    {
+      "epoch": 1.472270150357407,
+      "grad_norm": 0.15972352925273195,
+      "learning_rate": 5.157725847838702e-06,
+      "loss": 0.4639,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4732561005669214,
+      "grad_norm": 0.16660573318241859,
+      "learning_rate": 5.152556141111295e-06,
+      "loss": 0.4719,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4742420507764358,
+      "grad_norm": 0.18062616563596273,
+      "learning_rate": 5.147386271138916e-06,
+      "loss": 0.4885,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4752280009859502,
+      "grad_norm": 0.1659592283655134,
+      "learning_rate": 5.142216243453657e-06,
+      "loss": 0.458,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4762139511954646,
+      "grad_norm": 0.16673357568626188,
+      "learning_rate": 5.137046063587789e-06,
+      "loss": 0.4655,
+      "step": 1497
+    },
+    {
+      "epoch": 1.477199901404979,
+      "grad_norm": 0.16242794037498784,
+      "learning_rate": 5.131875737073736e-06,
+      "loss": 0.4653,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4781858516144935,
+      "grad_norm": 0.16336059472360087,
+      "learning_rate": 5.126705269444084e-06,
+      "loss": 0.4488,
+      "step": 1499
+    },
+    {
+      "epoch": 1.4791718018240079,
+      "grad_norm": 0.16619479701184342,
+      "learning_rate": 5.1215346662315705e-06,
+      "loss": 0.4681,
+      "step": 1500
+    },
+    {
+      "epoch": 1.4801577520335223,
+      "grad_norm": 0.15907452378843642,
+      "learning_rate": 5.116363932969074e-06,
+      "loss": 0.4701,
+      "step": 1501
+    },
+    {
+      "epoch": 1.4811437022430367,
+      "grad_norm": 0.16682593799018364,
+      "learning_rate": 5.111193075189617e-06,
+      "loss": 0.4856,
+      "step": 1502
+    },
+    {
+      "epoch": 1.4821296524525511,
+      "grad_norm": 0.1643805853475963,
+      "learning_rate": 5.106022098426351e-06,
+      "loss": 0.4595,
+      "step": 1503
+    },
+    {
+      "epoch": 1.4831156026620655,
+      "grad_norm": 0.15870679737925164,
+      "learning_rate": 5.100851008212557e-06,
+      "loss": 0.4582,
+      "step": 1504
+    },
+    {
+      "epoch": 1.48410155287158,
+      "grad_norm": 0.15789400323136946,
+      "learning_rate": 5.095679810081641e-06,
+      "loss": 0.4776,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4850875030810944,
+      "grad_norm": 0.15848384285661024,
+      "learning_rate": 5.090508509567115e-06,
+      "loss": 0.4697,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4860734532906088,
+      "grad_norm": 0.16505008193274287,
+      "learning_rate": 5.08533711220261e-06,
+      "loss": 0.4665,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4870594035001232,
+      "grad_norm": 0.17116248764476494,
+      "learning_rate": 5.080165623521854e-06,
+      "loss": 0.4541,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4880453537096376,
+      "grad_norm": 0.1541353248533124,
+      "learning_rate": 5.0749940490586795e-06,
+      "loss": 0.4711,
+      "step": 1509
+    },
+    {
+      "epoch": 1.489031303919152,
+      "grad_norm": 0.1623999670338064,
+      "learning_rate": 5.069822394347004e-06,
+      "loss": 0.4818,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4900172541286665,
+      "grad_norm": 0.32249731811742466,
+      "learning_rate": 5.064650664920834e-06,
+      "loss": 0.475,
+      "step": 1511
+    },
+    {
+      "epoch": 1.491003204338181,
+      "grad_norm": 0.16355643012459464,
+      "learning_rate": 5.059478866314255e-06,
+      "loss": 0.4768,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4919891545476953,
+      "grad_norm": 0.15994494636484732,
+      "learning_rate": 5.05430700406143e-06,
+      "loss": 0.4566,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4929751047572097,
+      "grad_norm": 0.16280872589015538,
+      "learning_rate": 5.049135083696585e-06,
+      "loss": 0.4881,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4939610549667242,
+      "grad_norm": 0.1644086039173946,
+      "learning_rate": 5.04396311075401e-06,
+      "loss": 0.4654,
+      "step": 1515
+    },
+    {
+      "epoch": 1.4949470051762386,
+      "grad_norm": 0.16504330201013406,
+      "learning_rate": 5.038791090768055e-06,
+      "loss": 0.458,
+      "step": 1516
+    },
+    {
+      "epoch": 1.495932955385753,
+      "grad_norm": 0.17997686772783875,
+      "learning_rate": 5.033619029273112e-06,
+      "loss": 0.4736,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4969189055952674,
+      "grad_norm": 0.15963908599115606,
+      "learning_rate": 5.0284469318036285e-06,
+      "loss": 0.4536,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4979048558047818,
+      "grad_norm": 0.15979259663425793,
+      "learning_rate": 5.023274803894079e-06,
+      "loss": 0.4619,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4988908060142963,
+      "grad_norm": 0.1565926166086123,
+      "learning_rate": 5.018102651078981e-06,
+      "loss": 0.4569,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4998767562238107,
+      "grad_norm": 0.16128313111834786,
+      "learning_rate": 5.012930478892869e-06,
+      "loss": 0.4893,
+      "step": 1521
+    },
+    {
+      "epoch": 1.500862706433325,
+      "grad_norm": 0.16573407567036658,
+      "learning_rate": 5.0077582928703065e-06,
+      "loss": 0.47,
+      "step": 1522
+    },
+    {
+      "epoch": 1.5018486566428395,
+      "grad_norm": 0.1587716773283982,
+      "learning_rate": 5.002586098545867e-06,
+      "loss": 0.4637,
+      "step": 1523
+    },
+    {
+      "epoch": 1.502834606852354,
+      "grad_norm": 0.15964197854300818,
+      "learning_rate": 4.997413901454134e-06,
+      "loss": 0.4519,
+      "step": 1524
+    },
+    {
+      "epoch": 1.5038205570618683,
+      "grad_norm": 0.15912646887006812,
+      "learning_rate": 4.9922417071296935e-06,
+      "loss": 0.4649,
+      "step": 1525
+    },
+    {
+      "epoch": 1.5048065072713828,
+      "grad_norm": 0.17354657537387477,
+      "learning_rate": 4.987069521107131e-06,
+      "loss": 0.4517,
+      "step": 1526
+    },
+    {
+      "epoch": 1.5057924574808972,
+      "grad_norm": 0.1669639920640026,
+      "learning_rate": 4.981897348921021e-06,
+      "loss": 0.4517,
+      "step": 1527
+    },
+    {
+      "epoch": 1.5067784076904116,
+      "grad_norm": 0.15981049076839046,
+      "learning_rate": 4.976725196105922e-06,
+      "loss": 0.4684,
+      "step": 1528
+    },
+    {
+      "epoch": 1.507764357899926,
+      "grad_norm": 0.16231859468822926,
+      "learning_rate": 4.971553068196373e-06,
+      "loss": 0.4502,
+      "step": 1529
+    },
+    {
+      "epoch": 1.5087503081094404,
+      "grad_norm": 0.17331787623465963,
+      "learning_rate": 4.966380970726889e-06,
+      "loss": 0.4512,
+      "step": 1530
+    },
+    {
+      "epoch": 1.5097362583189549,
+      "grad_norm": 0.1712976011319666,
+      "learning_rate": 4.961208909231946e-06,
+      "loss": 0.4866,
+      "step": 1531
+    },
+    {
+      "epoch": 1.5107222085284693,
+      "grad_norm": 0.16007715604197134,
+      "learning_rate": 4.956036889245991e-06,
+      "loss": 0.4321,
+      "step": 1532
+    },
+    {
+      "epoch": 1.5117081587379837,
+      "grad_norm": 0.16106511677415258,
+      "learning_rate": 4.950864916303417e-06,
+      "loss": 0.461,
+      "step": 1533
+    },
+    {
+      "epoch": 1.5126941089474981,
+      "grad_norm": 0.1607939029014845,
+      "learning_rate": 4.945692995938573e-06,
+      "loss": 0.462,
+      "step": 1534
+    },
+    {
+      "epoch": 1.5136800591570125,
+      "grad_norm": 0.16294803618833043,
+      "learning_rate": 4.940521133685746e-06,
+      "loss": 0.4719,
+      "step": 1535
+    },
+    {
+      "epoch": 1.514666009366527,
+      "grad_norm": 0.16300678600106153,
+      "learning_rate": 4.935349335079168e-06,
+      "loss": 0.4754,
+      "step": 1536
+    },
+    {
+      "epoch": 1.5156519595760414,
+      "grad_norm": 0.16337752099860722,
+      "learning_rate": 4.930177605652999e-06,
+      "loss": 0.4769,
+      "step": 1537
+    },
+    {
+      "epoch": 1.5166379097855558,
+      "grad_norm": 0.17190254222704324,
+      "learning_rate": 4.925005950941322e-06,
+      "loss": 0.4773,
+      "step": 1538
+    },
+    {
+      "epoch": 1.5176238599950702,
+      "grad_norm": 0.16075229340350197,
+      "learning_rate": 4.919834376478147e-06,
+      "loss": 0.4537,
+      "step": 1539
+    },
+    {
+      "epoch": 1.5186098102045846,
+      "grad_norm": 0.16430904130065005,
+      "learning_rate": 4.914662887797391e-06,
+      "loss": 0.4771,
+      "step": 1540
+    },
+    {
+      "epoch": 1.519595760414099,
+      "grad_norm": 0.1590528354097433,
+      "learning_rate": 4.909491490432886e-06,
+      "loss": 0.4659,
+      "step": 1541
+    },
+    {
+      "epoch": 1.5205817106236135,
+      "grad_norm": 0.16323440850911192,
+      "learning_rate": 4.904320189918362e-06,
+      "loss": 0.4754,
+      "step": 1542
+    },
+    {
+      "epoch": 1.5215676608331279,
+      "grad_norm": 0.16394665404460873,
+      "learning_rate": 4.899148991787444e-06,
+      "loss": 0.4664,
+      "step": 1543
+    },
+    {
+      "epoch": 1.5225536110426423,
+      "grad_norm": 0.16259969161433288,
+      "learning_rate": 4.893977901573651e-06,
+      "loss": 0.4695,
+      "step": 1544
+    },
+    {
+      "epoch": 1.5235395612521567,
+      "grad_norm": 0.1697390736728767,
+      "learning_rate": 4.888806924810385e-06,
+      "loss": 0.4807,
+      "step": 1545
+    },
+    {
+      "epoch": 1.5245255114616711,
+      "grad_norm": 0.16879485066467068,
+      "learning_rate": 4.883636067030927e-06,
+      "loss": 0.4772,
+      "step": 1546
+    },
+    {
+      "epoch": 1.5255114616711856,
+      "grad_norm": 0.16094250386353298,
+      "learning_rate": 4.878465333768432e-06,
+      "loss": 0.4771,
+      "step": 1547
+    },
+    {
+      "epoch": 1.5264974118807,
+      "grad_norm": 0.16603533601441273,
+      "learning_rate": 4.873294730555917e-06,
+      "loss": 0.4579,
+      "step": 1548
+    },
+    {
+      "epoch": 1.5274833620902144,
+      "grad_norm": 0.1662652078835514,
+      "learning_rate": 4.868124262926266e-06,
+      "loss": 0.4482,
+      "step": 1549
+    },
+    {
+      "epoch": 1.5284693122997288,
+      "grad_norm": 0.2053258025193838,
+      "learning_rate": 4.862953936412212e-06,
+      "loss": 0.4744,
+      "step": 1550
+    },
+    {
+      "epoch": 1.5294552625092432,
+      "grad_norm": 0.1610932388153637,
+      "learning_rate": 4.857783756546343e-06,
+      "loss": 0.4697,
+      "step": 1551
+    },
+    {
+      "epoch": 1.5304412127187577,
+      "grad_norm": 0.16323632481812023,
+      "learning_rate": 4.852613728861087e-06,
+      "loss": 0.4596,
+      "step": 1552
+    },
+    {
+      "epoch": 1.531427162928272,
+      "grad_norm": 0.15864551904706486,
+      "learning_rate": 4.847443858888707e-06,
+      "loss": 0.4642,
+      "step": 1553
+    },
+    {
+      "epoch": 1.5324131131377865,
+      "grad_norm": 0.16363552393779485,
+      "learning_rate": 4.842274152161298e-06,
+      "loss": 0.4665,
+      "step": 1554
+    },
+    {
+      "epoch": 1.533399063347301,
+      "grad_norm": 0.1606011248078566,
+      "learning_rate": 4.8371046142107865e-06,
+      "loss": 0.4547,
+      "step": 1555
+    },
+    {
+      "epoch": 1.5343850135568153,
+      "grad_norm": 0.16610436957840455,
+      "learning_rate": 4.831935250568911e-06,
+      "loss": 0.4789,
+      "step": 1556
+    },
+    {
+      "epoch": 1.5353709637663298,
+      "grad_norm": 0.16457826211107202,
+      "learning_rate": 4.826766066767228e-06,
+      "loss": 0.4577,
+      "step": 1557
+    },
+    {
+      "epoch": 1.5363569139758442,
+      "grad_norm": 0.1583944339333812,
+      "learning_rate": 4.821597068337097e-06,
+      "loss": 0.4686,
+      "step": 1558
+    },
+    {
+      "epoch": 1.5373428641853586,
+      "grad_norm": 0.2130665095886175,
+      "learning_rate": 4.816428260809682e-06,
+      "loss": 0.4815,
+      "step": 1559
+    },
+    {
+      "epoch": 1.538328814394873,
+      "grad_norm": 0.1717835073413018,
+      "learning_rate": 4.811259649715945e-06,
+      "loss": 0.478,
+      "step": 1560
+    },
+    {
+      "epoch": 1.5393147646043874,
+      "grad_norm": 0.16485264222155294,
+      "learning_rate": 4.806091240586633e-06,
+      "loss": 0.4796,
+      "step": 1561
+    },
+    {
+      "epoch": 1.5403007148139019,
+      "grad_norm": 0.16782791251902787,
+      "learning_rate": 4.800923038952282e-06,
+      "loss": 0.4606,
+      "step": 1562
+    },
+    {
+      "epoch": 1.5412866650234163,
+      "grad_norm": 0.15869786778702402,
+      "learning_rate": 4.795755050343199e-06,
+      "loss": 0.4613,
+      "step": 1563
+    },
+    {
+      "epoch": 1.5422726152329307,
+      "grad_norm": 0.16360053404141608,
+      "learning_rate": 4.790587280289472e-06,
+      "loss": 0.4677,
+      "step": 1564
+    },
+    {
+      "epoch": 1.543258565442445,
+      "grad_norm": 0.17053556390655156,
+      "learning_rate": 4.785419734320946e-06,
+      "loss": 0.4779,
+      "step": 1565
+    },
+    {
+      "epoch": 1.5442445156519595,
+      "grad_norm": 0.16345973492325022,
+      "learning_rate": 4.780252417967234e-06,
+      "loss": 0.4794,
+      "step": 1566
+    },
+    {
+      "epoch": 1.545230465861474,
+      "grad_norm": 0.1624815669558633,
+      "learning_rate": 4.775085336757699e-06,
+      "loss": 0.4815,
+      "step": 1567
+    },
+    {
+      "epoch": 1.5462164160709884,
+      "grad_norm": 0.16126412585156985,
+      "learning_rate": 4.7699184962214526e-06,
+      "loss": 0.4676,
+      "step": 1568
+    },
+    {
+      "epoch": 1.5472023662805028,
+      "grad_norm": 0.3017180723300519,
+      "learning_rate": 4.764751901887349e-06,
+      "loss": 0.4688,
+      "step": 1569
+    },
+    {
+      "epoch": 1.5481883164900172,
+      "grad_norm": 0.16701523355162504,
+      "learning_rate": 4.759585559283981e-06,
+      "loss": 0.4846,
+      "step": 1570
+    },
+    {
+      "epoch": 1.5491742666995316,
+      "grad_norm": 0.16211400835309955,
+      "learning_rate": 4.754419473939669e-06,
+      "loss": 0.4758,
+      "step": 1571
+    },
+    {
+      "epoch": 1.550160216909046,
+      "grad_norm": 0.1564366269152217,
+      "learning_rate": 4.7492536513824634e-06,
+      "loss": 0.4503,
+      "step": 1572
+    },
+    {
+      "epoch": 1.5511461671185605,
+      "grad_norm": 0.17540862590653875,
+      "learning_rate": 4.744088097140125e-06,
+      "loss": 0.4452,
+      "step": 1573
+    },
+    {
+      "epoch": 1.5521321173280749,
+      "grad_norm": 0.15707201548033853,
+      "learning_rate": 4.738922816740134e-06,
+      "loss": 0.4548,
+      "step": 1574
+    },
+    {
+      "epoch": 1.5531180675375893,
+      "grad_norm": 0.16604684393940009,
+      "learning_rate": 4.733757815709675e-06,
+      "loss": 0.4701,
+      "step": 1575
+    },
+    {
+      "epoch": 1.5541040177471037,
+      "grad_norm": 0.16027270475518166,
+      "learning_rate": 4.7285930995756355e-06,
+      "loss": 0.4637,
+      "step": 1576
+    },
+    {
+      "epoch": 1.5550899679566181,
+      "grad_norm": 0.16005286781324335,
+      "learning_rate": 4.7234286738645975e-06,
+      "loss": 0.4723,
+      "step": 1577
+    },
+    {
+      "epoch": 1.5560759181661328,
+      "grad_norm": 0.16517741964187063,
+      "learning_rate": 4.718264544102829e-06,
+      "loss": 0.4765,
+      "step": 1578
+    },
+    {
+      "epoch": 1.557061868375647,
+      "grad_norm": 0.17187802486105916,
+      "learning_rate": 4.713100715816287e-06,
+      "loss": 0.4653,
+      "step": 1579
+    },
+    {
+      "epoch": 1.5580478185851616,
+      "grad_norm": 0.15723170870463232,
+      "learning_rate": 4.707937194530602e-06,
+      "loss": 0.4597,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5590337687946758,
+      "grad_norm": 0.16060894388992308,
+      "learning_rate": 4.702773985771075e-06,
+      "loss": 0.4468,
+      "step": 1581
+    },
+    {
+      "epoch": 1.5600197190041905,
+      "grad_norm": 0.16123556569027944,
+      "learning_rate": 4.697611095062679e-06,
+      "loss": 0.4735,
+      "step": 1582
+    },
+    {
+      "epoch": 1.5610056692137046,
+      "grad_norm": 0.16803448436652849,
+      "learning_rate": 4.692448527930038e-06,
+      "loss": 0.4645,
+      "step": 1583
+    },
+    {
+      "epoch": 1.5619916194232193,
+      "grad_norm": 0.1597392579513729,
+      "learning_rate": 4.6872862898974345e-06,
+      "loss": 0.4621,
+      "step": 1584
+    },
+    {
+      "epoch": 1.5629775696327335,
+      "grad_norm": 0.15735379441805616,
+      "learning_rate": 4.6821243864888e-06,
+      "loss": 0.4709,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5639635198422481,
+      "grad_norm": 0.15404594214049994,
+      "learning_rate": 4.676962823227704e-06,
+      "loss": 0.4648,
+      "step": 1586
+    },
+    {
+      "epoch": 1.5649494700517623,
+      "grad_norm": 0.1615620632965602,
+      "learning_rate": 4.67180160563736e-06,
+      "loss": 0.4564,
+      "step": 1587
+    },
+    {
+      "epoch": 1.565935420261277,
+      "grad_norm": 0.16706931355044777,
+      "learning_rate": 4.666640739240596e-06,
+      "loss": 0.4577,
+      "step": 1588
+    },
+    {
+      "epoch": 1.5669213704707912,
+      "grad_norm": 0.16696669317402332,
+      "learning_rate": 4.661480229559882e-06,
+      "loss": 0.4842,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5679073206803058,
+      "grad_norm": 0.1579948637916331,
+      "learning_rate": 4.656320082117295e-06,
+      "loss": 0.4547,
+      "step": 1590
+    },
+    {
+      "epoch": 1.56889327088982,
+      "grad_norm": 0.15781583927294385,
+      "learning_rate": 4.6511603024345286e-06,
+      "loss": 0.4647,
+      "step": 1591
+    },
+    {
+      "epoch": 1.5698792210993346,
+      "grad_norm": 0.2129761427535933,
+      "learning_rate": 4.6460008960328834e-06,
+      "loss": 0.4487,
+      "step": 1592
+    },
+    {
+      "epoch": 1.5708651713088488,
+      "grad_norm": 0.2337698011010383,
+      "learning_rate": 4.640841868433257e-06,
+      "loss": 0.4575,
+      "step": 1593
+    },
+    {
+      "epoch": 1.5718511215183635,
+      "grad_norm": 0.16212957242189352,
+      "learning_rate": 4.635683225156142e-06,
+      "loss": 0.4696,
+      "step": 1594
+    },
+    {
+      "epoch": 1.5728370717278777,
+      "grad_norm": 0.16507879401280473,
+      "learning_rate": 4.630524971721626e-06,
+      "loss": 0.4648,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5738230219373923,
+      "grad_norm": 0.16957907647895618,
+      "learning_rate": 4.625367113649371e-06,
+      "loss": 0.4868,
+      "step": 1596
+    },
+    {
+      "epoch": 1.5748089721469065,
+      "grad_norm": 0.16317374423938397,
+      "learning_rate": 4.620209656458626e-06,
+      "loss": 0.4871,
+      "step": 1597
+    },
+    {
+      "epoch": 1.5757949223564212,
+      "grad_norm": 0.16220240581447787,
+      "learning_rate": 4.615052605668198e-06,
+      "loss": 0.4669,
+      "step": 1598
+    },
+    {
+      "epoch": 1.5767808725659354,
+      "grad_norm": 0.16935425806114593,
+      "learning_rate": 4.609895966796469e-06,
+      "loss": 0.4679,
+      "step": 1599
+    },
+    {
+      "epoch": 1.57776682277545,
+      "grad_norm": 0.16292461772554667,
+      "learning_rate": 4.604739745361377e-06,
+      "loss": 0.4421,
+      "step": 1600
+    },
+    {
+      "epoch": 1.5787527729849642,
+      "grad_norm": 0.1605822869671949,
+      "learning_rate": 4.599583946880415e-06,
+      "loss": 0.4593,
+      "step": 1601
+    },
+    {
+      "epoch": 1.5797387231944788,
+      "grad_norm": 0.16032696814199088,
+      "learning_rate": 4.594428576870622e-06,
+      "loss": 0.467,
+      "step": 1602
+    },
+    {
+      "epoch": 1.580724673403993,
+      "grad_norm": 0.16240637061318583,
+      "learning_rate": 4.589273640848575e-06,
+      "loss": 0.4717,
+      "step": 1603
+    },
+    {
+      "epoch": 1.5817106236135077,
+      "grad_norm": 0.16418809930496206,
+      "learning_rate": 4.584119144330394e-06,
+      "loss": 0.478,
+      "step": 1604
+    },
+    {
+      "epoch": 1.5826965738230219,
+      "grad_norm": 0.17631339108317104,
+      "learning_rate": 4.578965092831722e-06,
+      "loss": 0.4878,
+      "step": 1605
+    },
+    {
+      "epoch": 1.5836825240325365,
+      "grad_norm": 0.16036063679879956,
+      "learning_rate": 4.5738114918677315e-06,
+      "loss": 0.4712,
+      "step": 1606
+    },
+    {
+      "epoch": 1.5846684742420507,
+      "grad_norm": 0.1605868130676638,
+      "learning_rate": 4.568658346953109e-06,
+      "loss": 0.4926,
+      "step": 1607
+    },
+    {
+      "epoch": 1.5856544244515653,
+      "grad_norm": 0.16388430528523226,
+      "learning_rate": 4.563505663602054e-06,
+      "loss": 0.4574,
+      "step": 1608
+    },
+    {
+      "epoch": 1.5866403746610795,
+      "grad_norm": 0.16391718314552686,
+      "learning_rate": 4.558353447328271e-06,
+      "loss": 0.4617,
+      "step": 1609
+    },
+    {
+      "epoch": 1.5876263248705942,
+      "grad_norm": 0.16632160236460128,
+      "learning_rate": 4.553201703644966e-06,
+      "loss": 0.4643,
+      "step": 1610
+    },
+    {
+      "epoch": 1.5886122750801084,
+      "grad_norm": 0.17815327746175302,
+      "learning_rate": 4.54805043806484e-06,
+      "loss": 0.4697,
+      "step": 1611
+    },
+    {
+      "epoch": 1.589598225289623,
+      "grad_norm": 0.16042376911405104,
+      "learning_rate": 4.542899656100082e-06,
+      "loss": 0.4638,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5905841754991372,
+      "grad_norm": 0.16155445892241058,
+      "learning_rate": 4.5377493632623644e-06,
+      "loss": 0.4662,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5915701257086519,
+      "grad_norm": 0.16766169662221317,
+      "learning_rate": 4.532599565062831e-06,
+      "loss": 0.4596,
+      "step": 1614
+    },
+    {
+      "epoch": 1.592556075918166,
+      "grad_norm": 0.17423163991048965,
+      "learning_rate": 4.527450267012101e-06,
+      "loss": 0.4677,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5935420261276807,
+      "grad_norm": 0.15822696845412731,
+      "learning_rate": 4.52230147462026e-06,
+      "loss": 0.4441,
+      "step": 1616
+    },
+    {
+      "epoch": 1.594527976337195,
+      "grad_norm": 0.16065271742379106,
+      "learning_rate": 4.517153193396847e-06,
+      "loss": 0.4548,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5955139265467095,
+      "grad_norm": 0.15326255834855834,
+      "learning_rate": 4.5120054288508615e-06,
+      "loss": 0.4367,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5964998767562237,
+      "grad_norm": 0.16989129262402686,
+      "learning_rate": 4.506858186490743e-06,
+      "loss": 0.4661,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5974858269657384,
+      "grad_norm": 0.16496856927193837,
+      "learning_rate": 4.501711471824373e-06,
+      "loss": 0.4677,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5984717771752526,
+      "grad_norm": 0.16312540567729633,
+      "learning_rate": 4.496565290359072e-06,
+      "loss": 0.4702,
+      "step": 1621
+    },
+    {
+      "epoch": 1.5994577273847672,
+      "grad_norm": 0.15744378380707014,
+      "learning_rate": 4.49141964760159e-06,
+      "loss": 0.4421,
+      "step": 1622
+    },
+    {
+      "epoch": 1.6004436775942814,
+      "grad_norm": 0.16246223308682423,
+      "learning_rate": 4.486274549058097e-06,
+      "loss": 0.4628,
+      "step": 1623
+    },
+    {
+      "epoch": 1.601429627803796,
+      "grad_norm": 0.1626597411867869,
+      "learning_rate": 4.481130000234181e-06,
+      "loss": 0.4698,
+      "step": 1624
+    },
+    {
+      "epoch": 1.6024155780133102,
+      "grad_norm": 0.16303281734251007,
+      "learning_rate": 4.475986006634845e-06,
+      "loss": 0.4809,
+      "step": 1625
+    },
+    {
+      "epoch": 1.603401528222825,
+      "grad_norm": 0.15722472582946692,
+      "learning_rate": 4.470842573764497e-06,
+      "loss": 0.456,
+      "step": 1626
+    },
+    {
+      "epoch": 1.604387478432339,
+      "grad_norm": 0.16625605076459507,
+      "learning_rate": 4.465699707126941e-06,
+      "loss": 0.4429,
+      "step": 1627
+    },
+    {
+      "epoch": 1.6053734286418537,
+      "grad_norm": 0.18406674358801198,
+      "learning_rate": 4.460557412225382e-06,
+      "loss": 0.4699,
+      "step": 1628
+    },
+    {
+      "epoch": 1.606359378851368,
+      "grad_norm": 0.1721080150370503,
+      "learning_rate": 4.455415694562406e-06,
+      "loss": 0.4697,
+      "step": 1629
+    },
+    {
+      "epoch": 1.6073453290608826,
+      "grad_norm": 0.16948457739433334,
+      "learning_rate": 4.450274559639985e-06,
+      "loss": 0.4678,
+      "step": 1630
+    },
+    {
+      "epoch": 1.6083312792703968,
+      "grad_norm": 0.15883455112278896,
+      "learning_rate": 4.44513401295947e-06,
+      "loss": 0.4687,
+      "step": 1631
+    },
+    {
+      "epoch": 1.6093172294799114,
+      "grad_norm": 0.15632932325561036,
+      "learning_rate": 4.4399940600215755e-06,
+      "loss": 0.4517,
+      "step": 1632
+    },
+    {
+      "epoch": 1.6103031796894256,
+      "grad_norm": 0.16230052278603177,
+      "learning_rate": 4.434854706326391e-06,
+      "loss": 0.4729,
+      "step": 1633
+    },
+    {
+      "epoch": 1.6112891298989402,
+      "grad_norm": 0.15862028377618406,
+      "learning_rate": 4.42971595737335e-06,
+      "loss": 0.4564,
+      "step": 1634
+    },
+    {
+      "epoch": 1.6122750801084544,
+      "grad_norm": 0.16736418715930995,
+      "learning_rate": 4.424577818661255e-06,
+      "loss": 0.4691,
+      "step": 1635
+    },
+    {
+      "epoch": 1.613261030317969,
+      "grad_norm": 0.1711039646803351,
+      "learning_rate": 4.419440295688241e-06,
+      "loss": 0.4627,
+      "step": 1636
+    },
+    {
+      "epoch": 1.6142469805274833,
+      "grad_norm": 0.15903724979653044,
+      "learning_rate": 4.4143033939517975e-06,
+      "loss": 0.4729,
+      "step": 1637
+    },
+    {
+      "epoch": 1.615232930736998,
+      "grad_norm": 0.1607537226026667,
+      "learning_rate": 4.409167118948742e-06,
+      "loss": 0.4786,
+      "step": 1638
+    },
+    {
+      "epoch": 1.6162188809465121,
+      "grad_norm": 5.445659751986802,
+      "learning_rate": 4.404031476175218e-06,
+      "loss": 0.4643,
+      "step": 1639
+    },
+    {
+      "epoch": 1.6172048311560268,
+      "grad_norm": 0.16996628246894466,
+      "learning_rate": 4.398896471126698e-06,
+      "loss": 0.4655,
+      "step": 1640
+    },
+    {
+      "epoch": 1.618190781365541,
+      "grad_norm": 0.17309182559629868,
+      "learning_rate": 4.393762109297973e-06,
+      "loss": 0.4658,
+      "step": 1641
+    },
+    {
+      "epoch": 1.6191767315750556,
+      "grad_norm": 0.16078101698207184,
+      "learning_rate": 4.388628396183141e-06,
+      "loss": 0.4393,
+      "step": 1642
+    },
+    {
+      "epoch": 1.6201626817845698,
+      "grad_norm": 0.1629599532326369,
+      "learning_rate": 4.383495337275611e-06,
+      "loss": 0.4589,
+      "step": 1643
+    },
+    {
+      "epoch": 1.6211486319940844,
+      "grad_norm": 0.1550532532080627,
+      "learning_rate": 4.378362938068087e-06,
+      "loss": 0.457,
+      "step": 1644
+    },
+    {
+      "epoch": 1.6221345822035986,
+      "grad_norm": 0.16125937412625427,
+      "learning_rate": 4.3732312040525694e-06,
+      "loss": 0.4645,
+      "step": 1645
+    },
+    {
+      "epoch": 1.6231205324131133,
+      "grad_norm": 0.15873540121958585,
+      "learning_rate": 4.368100140720347e-06,
+      "loss": 0.4731,
+      "step": 1646
+    },
+    {
+      "epoch": 1.6241064826226275,
+      "grad_norm": 0.17053084904384064,
+      "learning_rate": 4.362969753561992e-06,
+      "loss": 0.4716,
+      "step": 1647
+    },
+    {
+      "epoch": 1.625092432832142,
+      "grad_norm": 0.16579599401772996,
+      "learning_rate": 4.357840048067351e-06,
+      "loss": 0.4747,
+      "step": 1648
+    },
+    {
+      "epoch": 1.6260783830416563,
+      "grad_norm": 0.18203093078322705,
+      "learning_rate": 4.352711029725539e-06,
+      "loss": 0.4582,
+      "step": 1649
+    },
+    {
+      "epoch": 1.627064333251171,
+      "grad_norm": 0.1687855099891574,
+      "learning_rate": 4.347582704024942e-06,
+      "loss": 0.4528,
+      "step": 1650
+    },
+    {
+      "epoch": 1.6280502834606851,
+      "grad_norm": 0.16436232071081422,
+      "learning_rate": 4.3424550764531995e-06,
+      "loss": 0.4677,
+      "step": 1651
+    },
+    {
+      "epoch": 1.6290362336701998,
+      "grad_norm": 0.17223817477890296,
+      "learning_rate": 4.337328152497206e-06,
+      "loss": 0.458,
+      "step": 1652
+    },
+    {
+      "epoch": 1.630022183879714,
+      "grad_norm": 0.16785666118136983,
+      "learning_rate": 4.332201937643107e-06,
+      "loss": 0.4709,
+      "step": 1653
+    },
+    {
+      "epoch": 1.6310081340892286,
+      "grad_norm": 0.16085219192688313,
+      "learning_rate": 4.3270764373762796e-06,
+      "loss": 0.4787,
+      "step": 1654
+    },
+    {
+      "epoch": 1.6319940842987428,
+      "grad_norm": 0.16254146140856732,
+      "learning_rate": 4.321951657181343e-06,
+      "loss": 0.4763,
+      "step": 1655
+    },
+    {
+      "epoch": 1.6329800345082575,
+      "grad_norm": 0.1598557744687256,
+      "learning_rate": 4.316827602542146e-06,
+      "loss": 0.4612,
+      "step": 1656
+    },
+    {
+      "epoch": 1.6339659847177717,
+      "grad_norm": 0.15979576055073055,
+      "learning_rate": 4.3117042789417586e-06,
+      "loss": 0.468,
+      "step": 1657
+    },
+    {
+      "epoch": 1.6349519349272863,
+      "grad_norm": 0.16708220010734257,
+      "learning_rate": 4.306581691862471e-06,
+      "loss": 0.4596,
+      "step": 1658
+    },
+    {
+      "epoch": 1.6359378851368005,
+      "grad_norm": 0.1600582187214164,
+      "learning_rate": 4.301459846785784e-06,
+      "loss": 0.4488,
+      "step": 1659
+    },
+    {
+      "epoch": 1.6369238353463151,
+      "grad_norm": 0.16661403488418633,
+      "learning_rate": 4.2963387491924015e-06,
+      "loss": 0.477,
+      "step": 1660
+    },
+    {
+      "epoch": 1.6379097855558293,
+      "grad_norm": 0.15859897062400877,
+      "learning_rate": 4.2912184045622325e-06,
+      "loss": 0.483,
+      "step": 1661
+    },
+    {
+      "epoch": 1.638895735765344,
+      "grad_norm": 0.16991212485166873,
+      "learning_rate": 4.2860988183743785e-06,
+      "loss": 0.4574,
+      "step": 1662
+    },
+    {
+      "epoch": 1.6398816859748582,
+      "grad_norm": 0.16312737419856346,
+      "learning_rate": 4.280979996107129e-06,
+      "loss": 0.4533,
+      "step": 1663
+    },
+    {
+      "epoch": 1.6408676361843728,
+      "grad_norm": 0.15925356100580668,
+      "learning_rate": 4.275861943237953e-06,
+      "loss": 0.4583,
+      "step": 1664
+    },
+    {
+      "epoch": 1.641853586393887,
+      "grad_norm": 0.16127114186632893,
+      "learning_rate": 4.270744665243504e-06,
+      "loss": 0.4606,
+      "step": 1665
+    },
+    {
+      "epoch": 1.6428395366034017,
+      "grad_norm": 0.16111245788776873,
+      "learning_rate": 4.265628167599599e-06,
+      "loss": 0.4588,
+      "step": 1666
+    },
+    {
+      "epoch": 1.6438254868129158,
+      "grad_norm": 0.16822860303261428,
+      "learning_rate": 4.260512455781221e-06,
+      "loss": 0.4807,
+      "step": 1667
+    },
+    {
+      "epoch": 1.6448114370224305,
+      "grad_norm": 0.15926862917951995,
+      "learning_rate": 4.255397535262518e-06,
+      "loss": 0.4551,
+      "step": 1668
+    },
+    {
+      "epoch": 1.6457973872319447,
+      "grad_norm": 0.1592182907342588,
+      "learning_rate": 4.250283411516784e-06,
+      "loss": 0.4562,
+      "step": 1669
+    },
+    {
+      "epoch": 1.6467833374414593,
+      "grad_norm": 0.16578470661790773,
+      "learning_rate": 4.245170090016463e-06,
+      "loss": 0.46,
+      "step": 1670
+    },
+    {
+      "epoch": 1.6477692876509735,
+      "grad_norm": 0.16954867487540962,
+      "learning_rate": 4.240057576233142e-06,
+      "loss": 0.4662,
+      "step": 1671
+    },
+    {
+      "epoch": 1.6487552378604882,
+      "grad_norm": 0.16085933560840476,
+      "learning_rate": 4.234945875637543e-06,
+      "loss": 0.4656,
+      "step": 1672
+    },
+    {
+      "epoch": 1.6497411880700024,
+      "grad_norm": 0.15603353053978614,
+      "learning_rate": 4.229834993699518e-06,
+      "loss": 0.4515,
+      "step": 1673
+    },
+    {
+      "epoch": 1.650727138279517,
+      "grad_norm": 0.1600911760925506,
+      "learning_rate": 4.224724935888039e-06,
+      "loss": 0.4741,
+      "step": 1674
+    },
+    {
+      "epoch": 1.6517130884890312,
+      "grad_norm": 0.15705474690537036,
+      "learning_rate": 4.219615707671204e-06,
+      "loss": 0.454,
+      "step": 1675
+    },
+    {
+      "epoch": 1.6526990386985458,
+      "grad_norm": 0.15647312431168459,
+      "learning_rate": 4.214507314516214e-06,
+      "loss": 0.4631,
+      "step": 1676
+    },
+    {
+      "epoch": 1.65368498890806,
+      "grad_norm": 0.1686109701449764,
+      "learning_rate": 4.2093997618893865e-06,
+      "loss": 0.4626,
+      "step": 1677
+    },
+    {
+      "epoch": 1.6546709391175747,
+      "grad_norm": 0.16045455935162395,
+      "learning_rate": 4.204293055256131e-06,
+      "loss": 0.4718,
+      "step": 1678
+    },
+    {
+      "epoch": 1.6556568893270889,
+      "grad_norm": 0.15497677791978665,
+      "learning_rate": 4.1991872000809566e-06,
+      "loss": 0.4706,
+      "step": 1679
+    },
+    {
+      "epoch": 1.6566428395366035,
+      "grad_norm": 0.15549738678280184,
+      "learning_rate": 4.194082201827458e-06,
+      "loss": 0.4404,
+      "step": 1680
+    },
+    {
+      "epoch": 1.6576287897461177,
+      "grad_norm": 0.15676194115412878,
+      "learning_rate": 4.1889780659583165e-06,
+      "loss": 0.46,
+      "step": 1681
+    },
+    {
+      "epoch": 1.6586147399556324,
+      "grad_norm": 0.1596892622244796,
+      "learning_rate": 4.183874797935286e-06,
+      "loss": 0.4521,
+      "step": 1682
+    },
+    {
+      "epoch": 1.6596006901651466,
+      "grad_norm": 0.1597247735805591,
+      "learning_rate": 4.1787724032192e-06,
+      "loss": 0.4593,
+      "step": 1683
+    },
+    {
+      "epoch": 1.6605866403746612,
+      "grad_norm": 0.16019979591444788,
+      "learning_rate": 4.173670887269946e-06,
+      "loss": 0.4445,
+      "step": 1684
+    },
+    {
+      "epoch": 1.6615725905841754,
+      "grad_norm": 0.15940742116423462,
+      "learning_rate": 4.1685702555464815e-06,
+      "loss": 0.4675,
+      "step": 1685
+    },
+    {
+      "epoch": 1.66255854079369,
+      "grad_norm": 0.15907571330710468,
+      "learning_rate": 4.16347051350681e-06,
+      "loss": 0.46,
+      "step": 1686
+    },
+    {
+      "epoch": 1.6635444910032042,
+      "grad_norm": 0.1584986180064822,
+      "learning_rate": 4.1583716666079894e-06,
+      "loss": 0.4704,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6645304412127189,
+      "grad_norm": 0.161288876602103,
+      "learning_rate": 4.153273720306115e-06,
+      "loss": 0.4654,
+      "step": 1688
+    },
+    {
+      "epoch": 1.665516391422233,
+      "grad_norm": 0.16426213294525213,
+      "learning_rate": 4.148176680056323e-06,
+      "loss": 0.459,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6665023416317477,
+      "grad_norm": 0.15770348258781552,
+      "learning_rate": 4.143080551312775e-06,
+      "loss": 0.4648,
+      "step": 1690
+    },
+    {
+      "epoch": 1.667488291841262,
+      "grad_norm": 0.1624674809972469,
+      "learning_rate": 4.137985339528658e-06,
+      "loss": 0.4733,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6684742420507765,
+      "grad_norm": 0.16664291687162525,
+      "learning_rate": 4.132891050156183e-06,
+      "loss": 0.4672,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6694601922602907,
+      "grad_norm": 0.15942602998198058,
+      "learning_rate": 4.127797688646568e-06,
+      "loss": 0.4798,
+      "step": 1693
+    },
+    {
+      "epoch": 1.6704461424698054,
+      "grad_norm": 0.15683100493840488,
+      "learning_rate": 4.1227052604500425e-06,
+      "loss": 0.4522,
+      "step": 1694
+    },
+    {
+      "epoch": 1.6714320926793196,
+      "grad_norm": 0.1612447220048338,
+      "learning_rate": 4.117613771015831e-06,
+      "loss": 0.4655,
+      "step": 1695
+    },
+    {
+      "epoch": 1.6724180428888342,
+      "grad_norm": 0.16599298633234355,
+      "learning_rate": 4.112523225792162e-06,
+      "loss": 0.4631,
+      "step": 1696
+    },
+    {
+      "epoch": 1.6734039930983484,
+      "grad_norm": 0.15757793172827048,
+      "learning_rate": 4.107433630226247e-06,
+      "loss": 0.4472,
+      "step": 1697
+    },
+    {
+      "epoch": 1.674389943307863,
+      "grad_norm": 0.16235656077954422,
+      "learning_rate": 4.102344989764285e-06,
+      "loss": 0.4808,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6753758935173773,
+      "grad_norm": 0.15966798991344844,
+      "learning_rate": 4.097257309851452e-06,
+      "loss": 0.4677,
+      "step": 1699
+    },
+    {
+      "epoch": 1.676361843726892,
+      "grad_norm": 0.16022835814471756,
+      "learning_rate": 4.092170595931893e-06,
+      "loss": 0.484,
+      "step": 1700
+    },
+    {
+      "epoch": 1.677347793936406,
+      "grad_norm": 0.15971748320658116,
+      "learning_rate": 4.0870848534487236e-06,
+      "loss": 0.4556,
+      "step": 1701
+    },
+    {
+      "epoch": 1.6783337441459207,
+      "grad_norm": 0.166482079806644,
+      "learning_rate": 4.082000087844019e-06,
+      "loss": 0.4644,
+      "step": 1702
+    },
+    {
+      "epoch": 1.679319694355435,
+      "grad_norm": 0.1583529523684524,
+      "learning_rate": 4.076916304558807e-06,
+      "loss": 0.4585,
+      "step": 1703
+    },
+    {
+      "epoch": 1.6803056445649496,
+      "grad_norm": 0.16078954728983477,
+      "learning_rate": 4.07183350903307e-06,
+      "loss": 0.459,
+      "step": 1704
+    },
+    {
+      "epoch": 1.6812915947744638,
+      "grad_norm": 0.16189131493190925,
+      "learning_rate": 4.066751706705723e-06,
+      "loss": 0.4558,
+      "step": 1705
+    },
+    {
+      "epoch": 1.6822775449839784,
+      "grad_norm": 0.1610716395237498,
+      "learning_rate": 4.061670903014629e-06,
+      "loss": 0.4496,
+      "step": 1706
+    },
+    {
+      "epoch": 1.6832634951934926,
+      "grad_norm": 0.16004742786983514,
+      "learning_rate": 4.056591103396573e-06,
+      "loss": 0.4647,
+      "step": 1707
+    },
+    {
+      "epoch": 1.6842494454030073,
+      "grad_norm": 0.16061135639699073,
+      "learning_rate": 4.051512313287276e-06,
+      "loss": 0.4644,
+      "step": 1708
+    },
+    {
+      "epoch": 1.6852353956125214,
+      "grad_norm": 0.15916518626370224,
+      "learning_rate": 4.04643453812137e-06,
+      "loss": 0.4682,
+      "step": 1709
+    },
+    {
+      "epoch": 1.686221345822036,
+      "grad_norm": 0.16356830805591838,
+      "learning_rate": 4.041357783332403e-06,
+      "loss": 0.4627,
+      "step": 1710
+    },
+    {
+      "epoch": 1.6872072960315503,
+      "grad_norm": 0.1659003580908028,
+      "learning_rate": 4.036282054352833e-06,
+      "loss": 0.462,
+      "step": 1711
+    },
+    {
+      "epoch": 1.688193246241065,
+      "grad_norm": 0.16266201853138954,
+      "learning_rate": 4.031207356614022e-06,
+      "loss": 0.4461,
+      "step": 1712
+    },
+    {
+      "epoch": 1.6891791964505791,
+      "grad_norm": 0.1641383969266073,
+      "learning_rate": 4.026133695546223e-06,
+      "loss": 0.4534,
+      "step": 1713
+    },
+    {
+      "epoch": 1.6901651466600938,
+      "grad_norm": 0.1546939101720392,
+      "learning_rate": 4.021061076578585e-06,
+      "loss": 0.4623,
+      "step": 1714
+    },
+    {
+      "epoch": 1.691151096869608,
+      "grad_norm": 0.1609467801963117,
+      "learning_rate": 4.015989505139137e-06,
+      "loss": 0.4575,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6921370470791226,
+      "grad_norm": 0.2973852373889811,
+      "learning_rate": 4.0109189866547896e-06,
+      "loss": 0.4637,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6931229972886368,
+      "grad_norm": 0.1649424082175255,
+      "learning_rate": 4.00584952655133e-06,
+      "loss": 0.4492,
+      "step": 1717
+    },
+    {
+      "epoch": 1.6941089474981514,
+      "grad_norm": 0.1974452838131342,
+      "learning_rate": 4.000781130253406e-06,
+      "loss": 0.4392,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6950948977076656,
+      "grad_norm": 0.15973190521591996,
+      "learning_rate": 3.995713803184535e-06,
+      "loss": 0.4795,
+      "step": 1719
+    },
+    {
+      "epoch": 1.6960808479171803,
+      "grad_norm": 0.1625275330726248,
+      "learning_rate": 3.99064755076708e-06,
+      "loss": 0.4526,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6970667981266945,
+      "grad_norm": 0.16264247139181023,
+      "learning_rate": 3.985582378422264e-06,
+      "loss": 0.4686,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6980527483362091,
+      "grad_norm": 0.16762631815091714,
+      "learning_rate": 3.980518291570148e-06,
+      "loss": 0.4697,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6990386985457233,
+      "grad_norm": 0.1592486161979069,
+      "learning_rate": 3.9754552956296365e-06,
+      "loss": 0.46,
+      "step": 1723
+    },
+    {
+      "epoch": 1.700024648755238,
+      "grad_norm": 0.1626241317549875,
+      "learning_rate": 3.970393396018462e-06,
+      "loss": 0.4616,
+      "step": 1724
+    },
+    {
+      "epoch": 1.7010105989647522,
+      "grad_norm": 0.15743672584041096,
+      "learning_rate": 3.965332598153186e-06,
+      "loss": 0.4648,
+      "step": 1725
+    },
+    {
+      "epoch": 1.7019965491742668,
+      "grad_norm": 0.16591668834389786,
+      "learning_rate": 3.9602729074491884e-06,
+      "loss": 0.4635,
+      "step": 1726
+    },
+    {
+      "epoch": 1.702982499383781,
+      "grad_norm": 0.16505167909686655,
+      "learning_rate": 3.955214329320671e-06,
+      "loss": 0.4673,
+      "step": 1727
+    },
+    {
+      "epoch": 1.7039684495932956,
+      "grad_norm": 0.15595137138501824,
+      "learning_rate": 3.950156869180637e-06,
+      "loss": 0.4507,
+      "step": 1728
+    },
+    {
+      "epoch": 1.7049543998028098,
+      "grad_norm": 0.16523673259698257,
+      "learning_rate": 3.9451005324409e-06,
+      "loss": 0.4783,
+      "step": 1729
+    },
+    {
+      "epoch": 1.7059403500123245,
+      "grad_norm": 0.1526757287169731,
+      "learning_rate": 3.940045324512066e-06,
+      "loss": 0.4521,
+      "step": 1730
+    },
+    {
+      "epoch": 1.7069263002218387,
+      "grad_norm": 0.15949669042895306,
+      "learning_rate": 3.934991250803537e-06,
+      "loss": 0.4742,
+      "step": 1731
+    },
+    {
+      "epoch": 1.7079122504313533,
+      "grad_norm": 0.16086159798759384,
+      "learning_rate": 3.929938316723499e-06,
+      "loss": 0.4552,
+      "step": 1732
+    },
+    {
+      "epoch": 1.7088982006408675,
+      "grad_norm": 0.16265387416782154,
+      "learning_rate": 3.924886527678921e-06,
+      "loss": 0.4623,
+      "step": 1733
+    },
+    {
+      "epoch": 1.7098841508503821,
+      "grad_norm": 0.15920090990234076,
+      "learning_rate": 3.919835889075545e-06,
+      "loss": 0.4549,
+      "step": 1734
+    },
+    {
+      "epoch": 1.7108701010598963,
+      "grad_norm": 0.15919083855836072,
+      "learning_rate": 3.914786406317879e-06,
+      "loss": 0.4507,
+      "step": 1735
+    },
+    {
+      "epoch": 1.711856051269411,
+      "grad_norm": 0.15985246069413103,
+      "learning_rate": 3.909738084809201e-06,
+      "loss": 0.4652,
+      "step": 1736
+    },
+    {
+      "epoch": 1.7128420014789252,
+      "grad_norm": 0.1625145221557171,
+      "learning_rate": 3.90469092995154e-06,
+      "loss": 0.4774,
+      "step": 1737
+    },
+    {
+      "epoch": 1.7138279516884398,
+      "grad_norm": 0.16428618121894853,
+      "learning_rate": 3.8996449471456825e-06,
+      "loss": 0.4798,
+      "step": 1738
+    },
+    {
+      "epoch": 1.714813901897954,
+      "grad_norm": 0.15946513268958273,
+      "learning_rate": 3.894600141791156e-06,
+      "loss": 0.4737,
+      "step": 1739
+    },
+    {
+      "epoch": 1.7157998521074687,
+      "grad_norm": 0.15438322650671502,
+      "learning_rate": 3.88955651928623e-06,
+      "loss": 0.4385,
+      "step": 1740
+    },
+    {
+      "epoch": 1.7167858023169829,
+      "grad_norm": 0.16197537413811997,
+      "learning_rate": 3.884514085027905e-06,
+      "loss": 0.4584,
+      "step": 1741
+    },
+    {
+      "epoch": 1.7177717525264975,
+      "grad_norm": 0.1591918416840697,
+      "learning_rate": 3.879472844411917e-06,
+      "loss": 0.4707,
+      "step": 1742
+    },
+    {
+      "epoch": 1.7187577027360117,
+      "grad_norm": 0.16316356543524038,
+      "learning_rate": 3.874432802832718e-06,
+      "loss": 0.4526,
+      "step": 1743
+    },
+    {
+      "epoch": 1.7197436529455263,
+      "grad_norm": 0.15830112702696214,
+      "learning_rate": 3.869393965683484e-06,
+      "loss": 0.4372,
+      "step": 1744
+    },
+    {
+      "epoch": 1.7207296031550405,
+      "grad_norm": 0.15868936240516976,
+      "learning_rate": 3.864356338356092e-06,
+      "loss": 0.4676,
+      "step": 1745
+    },
+    {
+      "epoch": 1.7217155533645552,
+      "grad_norm": 0.16180924305285438,
+      "learning_rate": 3.8593199262411335e-06,
+      "loss": 0.4598,
+      "step": 1746
+    },
+    {
+      "epoch": 1.7227015035740694,
+      "grad_norm": 0.16499470407397807,
+      "learning_rate": 3.854284734727895e-06,
+      "loss": 0.4629,
+      "step": 1747
+    },
+    {
+      "epoch": 1.723687453783584,
+      "grad_norm": 0.15922359738365144,
+      "learning_rate": 3.84925076920436e-06,
+      "loss": 0.4579,
+      "step": 1748
+    },
+    {
+      "epoch": 1.7246734039930982,
+      "grad_norm": 0.1666424462570064,
+      "learning_rate": 3.8442180350571974e-06,
+      "loss": 0.4675,
+      "step": 1749
+    },
+    {
+      "epoch": 1.7256593542026128,
+      "grad_norm": 0.16124955431233737,
+      "learning_rate": 3.839186537671758e-06,
+      "loss": 0.4643,
+      "step": 1750
+    },
+    {
+      "epoch": 1.726645304412127,
+      "grad_norm": 0.1677657662261242,
+      "learning_rate": 3.8341562824320724e-06,
+      "loss": 0.4664,
+      "step": 1751
+    },
+    {
+      "epoch": 1.7276312546216417,
+      "grad_norm": 0.15804633567061305,
+      "learning_rate": 3.829127274720841e-06,
+      "loss": 0.4466,
+      "step": 1752
+    },
+    {
+      "epoch": 1.7286172048311559,
+      "grad_norm": 0.15774444235500484,
+      "learning_rate": 3.8240995199194255e-06,
+      "loss": 0.4787,
+      "step": 1753
+    },
+    {
+      "epoch": 1.7296031550406705,
+      "grad_norm": 0.1547158604668105,
+      "learning_rate": 3.819073023407854e-06,
+      "loss": 0.4674,
+      "step": 1754
+    },
+    {
+      "epoch": 1.7305891052501847,
+      "grad_norm": 0.15481840956469606,
+      "learning_rate": 3.8140477905648e-06,
+      "loss": 0.4562,
+      "step": 1755
+    },
+    {
+      "epoch": 1.7315750554596994,
+      "grad_norm": 0.15551552639232807,
+      "learning_rate": 3.80902382676759e-06,
+      "loss": 0.4487,
+      "step": 1756
+    },
+    {
+      "epoch": 1.7325610056692136,
+      "grad_norm": 0.1594096173164566,
+      "learning_rate": 3.8040011373921925e-06,
+      "loss": 0.4629,
+      "step": 1757
+    },
+    {
+      "epoch": 1.7335469558787282,
+      "grad_norm": 0.1686421203994268,
+      "learning_rate": 3.798979727813211e-06,
+      "loss": 0.4721,
+      "step": 1758
+    },
+    {
+      "epoch": 1.7345329060882424,
+      "grad_norm": 0.1701916542255308,
+      "learning_rate": 3.7939596034038807e-06,
+      "loss": 0.4693,
+      "step": 1759
+    },
+    {
+      "epoch": 1.735518856297757,
+      "grad_norm": 0.1604063653101608,
+      "learning_rate": 3.7889407695360565e-06,
+      "loss": 0.4689,
+      "step": 1760
+    },
+    {
+      "epoch": 1.7365048065072712,
+      "grad_norm": 0.15803387279080147,
+      "learning_rate": 3.78392323158022e-06,
+      "loss": 0.4514,
+      "step": 1761
+    },
+    {
+      "epoch": 1.7374907567167859,
+      "grad_norm": 0.16542809413251114,
+      "learning_rate": 3.77890699490546e-06,
+      "loss": 0.4697,
+      "step": 1762
+    },
+    {
+      "epoch": 1.7384767069263,
+      "grad_norm": 0.16354444721337727,
+      "learning_rate": 3.7738920648794785e-06,
+      "loss": 0.4756,
+      "step": 1763
+    },
+    {
+      "epoch": 1.7394626571358147,
+      "grad_norm": 0.1669387947207977,
+      "learning_rate": 3.768878446868576e-06,
+      "loss": 0.4652,
+      "step": 1764
+    },
+    {
+      "epoch": 1.740448607345329,
+      "grad_norm": 0.15513379263677446,
+      "learning_rate": 3.7638661462376464e-06,
+      "loss": 0.4514,
+      "step": 1765
+    },
+    {
+      "epoch": 1.7414345575548436,
+      "grad_norm": 0.15457732346104522,
+      "learning_rate": 3.7588551683501767e-06,
+      "loss": 0.4557,
+      "step": 1766
+    },
+    {
+      "epoch": 1.7424205077643578,
+      "grad_norm": 0.15684459051807895,
+      "learning_rate": 3.753845518568241e-06,
+      "loss": 0.4675,
+      "step": 1767
+    },
+    {
+      "epoch": 1.7434064579738724,
+      "grad_norm": 0.1643668072910157,
+      "learning_rate": 3.748837202252488e-06,
+      "loss": 0.4613,
+      "step": 1768
+    },
+    {
+      "epoch": 1.7443924081833866,
+      "grad_norm": 0.16720662992421156,
+      "learning_rate": 3.7438302247621433e-06,
+      "loss": 0.4556,
+      "step": 1769
+    },
+    {
+      "epoch": 1.7453783583929012,
+      "grad_norm": 0.16017914106674547,
+      "learning_rate": 3.738824591454996e-06,
+      "loss": 0.4753,
+      "step": 1770
+    },
+    {
+      "epoch": 1.7463643086024154,
+      "grad_norm": 0.16304058666183993,
+      "learning_rate": 3.733820307687398e-06,
+      "loss": 0.4606,
+      "step": 1771
+    },
+    {
+      "epoch": 1.74735025881193,
+      "grad_norm": 0.15879668350572085,
+      "learning_rate": 3.7288173788142586e-06,
+      "loss": 0.4689,
+      "step": 1772
+    },
+    {
+      "epoch": 1.7483362090214443,
+      "grad_norm": 0.16213042148565254,
+      "learning_rate": 3.7238158101890376e-06,
+      "loss": 0.4811,
+      "step": 1773
+    },
+    {
+      "epoch": 1.749322159230959,
+      "grad_norm": 0.162055828243607,
+      "learning_rate": 3.718815607163736e-06,
+      "loss": 0.4846,
+      "step": 1774
+    },
+    {
+      "epoch": 1.750308109440473,
+      "grad_norm": 0.1625855396752446,
+      "learning_rate": 3.7138167750888985e-06,
+      "loss": 0.4623,
+      "step": 1775
+    },
+    {
+      "epoch": 1.7512940596499877,
+      "grad_norm": 0.15526605487307704,
+      "learning_rate": 3.708819319313597e-06,
+      "loss": 0.4621,
+      "step": 1776
+    },
+    {
+      "epoch": 1.752280009859502,
+      "grad_norm": 0.16113481794568882,
+      "learning_rate": 3.703823245185434e-06,
+      "loss": 0.4538,
+      "step": 1777
+    },
+    {
+      "epoch": 1.7532659600690166,
+      "grad_norm": 0.16413413164604834,
+      "learning_rate": 3.6988285580505345e-06,
+      "loss": 0.4737,
+      "step": 1778
+    },
+    {
+      "epoch": 1.7542519102785308,
+      "grad_norm": 0.15714016554624696,
+      "learning_rate": 3.693835263253538e-06,
+      "loss": 0.464,
+      "step": 1779
+    },
+    {
+      "epoch": 1.7552378604880454,
+      "grad_norm": 0.15705356629935893,
+      "learning_rate": 3.6888433661375934e-06,
+      "loss": 0.458,
+      "step": 1780
+    },
+    {
+      "epoch": 1.7562238106975596,
+      "grad_norm": 0.16381244221833122,
+      "learning_rate": 3.683852872044353e-06,
+      "loss": 0.467,
+      "step": 1781
+    },
+    {
+      "epoch": 1.7572097609070743,
+      "grad_norm": 0.16170181326889393,
+      "learning_rate": 3.6788637863139716e-06,
+      "loss": 0.4634,
+      "step": 1782
+    },
+    {
+      "epoch": 1.7581957111165885,
+      "grad_norm": 0.1843592684562403,
+      "learning_rate": 3.673876114285093e-06,
+      "loss": 0.4656,
+      "step": 1783
+    },
+    {
+      "epoch": 1.759181661326103,
+      "grad_norm": 0.15988105544370507,
+      "learning_rate": 3.668889861294852e-06,
+      "loss": 0.4761,
+      "step": 1784
+    },
+    {
+      "epoch": 1.7601676115356173,
+      "grad_norm": 0.16154518889017053,
+      "learning_rate": 3.6639050326788637e-06,
+      "loss": 0.4482,
+      "step": 1785
+    },
+    {
+      "epoch": 1.761153561745132,
+      "grad_norm": 0.16842499238085878,
+      "learning_rate": 3.6589216337712153e-06,
+      "loss": 0.4444,
+      "step": 1786
+    },
+    {
+      "epoch": 1.7621395119546461,
+      "grad_norm": 0.15648832899751022,
+      "learning_rate": 3.653939669904468e-06,
+      "loss": 0.4564,
+      "step": 1787
+    },
+    {
+      "epoch": 1.7631254621641608,
+      "grad_norm": 0.1561238377660689,
+      "learning_rate": 3.6489591464096475e-06,
+      "loss": 0.4554,
+      "step": 1788
+    },
+    {
+      "epoch": 1.764111412373675,
+      "grad_norm": 0.15793615530774216,
+      "learning_rate": 3.6439800686162354e-06,
+      "loss": 0.481,
+      "step": 1789
+    },
+    {
+      "epoch": 1.7650973625831896,
+      "grad_norm": 0.16722508169816377,
+      "learning_rate": 3.639002441852173e-06,
+      "loss": 0.4704,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7660833127927038,
+      "grad_norm": 0.15932686815960895,
+      "learning_rate": 3.634026271443837e-06,
+      "loss": 0.4488,
+      "step": 1791
+    },
+    {
+      "epoch": 1.7670692630022184,
+      "grad_norm": 0.16354950036583038,
+      "learning_rate": 3.629051562716058e-06,
+      "loss": 0.4701,
+      "step": 1792
+    },
+    {
+      "epoch": 1.7680552132117326,
+      "grad_norm": 0.15870969754085226,
+      "learning_rate": 3.624078320992094e-06,
+      "loss": 0.4509,
+      "step": 1793
+    },
+    {
+      "epoch": 1.7690411634212473,
+      "grad_norm": 0.16063785348184123,
+      "learning_rate": 3.6191065515936387e-06,
+      "loss": 0.4533,
+      "step": 1794
+    },
+    {
+      "epoch": 1.7700271136307615,
+      "grad_norm": 0.15752721509162096,
+      "learning_rate": 3.6141362598408087e-06,
+      "loss": 0.4633,
+      "step": 1795
+    },
+    {
+      "epoch": 1.7710130638402761,
+      "grad_norm": 0.15630592366386475,
+      "learning_rate": 3.609167451052135e-06,
+      "loss": 0.4639,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7719990140497903,
+      "grad_norm": 0.15733705201561435,
+      "learning_rate": 3.6042001305445693e-06,
+      "loss": 0.4666,
+      "step": 1797
+    },
+    {
+      "epoch": 1.772984964259305,
+      "grad_norm": 0.15860971912053975,
+      "learning_rate": 3.5992343036334653e-06,
+      "loss": 0.4607,
+      "step": 1798
+    },
+    {
+      "epoch": 1.7739709144688192,
+      "grad_norm": 0.16061398006664077,
+      "learning_rate": 3.5942699756325795e-06,
+      "loss": 0.4715,
+      "step": 1799
+    },
+    {
+      "epoch": 1.7749568646783338,
+      "grad_norm": 0.15145860278327483,
+      "learning_rate": 3.5893071518540683e-06,
+      "loss": 0.4335,
+      "step": 1800
+    },
+    {
+      "epoch": 1.775942814887848,
+      "grad_norm": 0.1592736663241254,
+      "learning_rate": 3.5843458376084715e-06,
+      "loss": 0.4658,
+      "step": 1801
+    },
+    {
+      "epoch": 1.7769287650973626,
+      "grad_norm": 0.16288409017228925,
+      "learning_rate": 3.5793860382047185e-06,
+      "loss": 0.4924,
+      "step": 1802
+    },
+    {
+      "epoch": 1.777914715306877,
+      "grad_norm": 0.1702474657072719,
+      "learning_rate": 3.5744277589501174e-06,
+      "loss": 0.4745,
+      "step": 1803
+    },
+    {
+      "epoch": 1.7789006655163915,
+      "grad_norm": 0.22498696327184153,
+      "learning_rate": 3.569471005150349e-06,
+      "loss": 0.4575,
+      "step": 1804
+    },
+    {
+      "epoch": 1.779886615725906,
+      "grad_norm": 0.15754237349064829,
+      "learning_rate": 3.5645157821094623e-06,
+      "loss": 0.4635,
+      "step": 1805
+    },
+    {
+      "epoch": 1.7808725659354203,
+      "grad_norm": 0.16167295071553067,
+      "learning_rate": 3.5595620951298637e-06,
+      "loss": 0.4728,
+      "step": 1806
+    },
+    {
+      "epoch": 1.7818585161449347,
+      "grad_norm": 0.1543577728513248,
+      "learning_rate": 3.554609949512324e-06,
+      "loss": 0.4512,
+      "step": 1807
+    },
+    {
+      "epoch": 1.7828444663544492,
+      "grad_norm": 0.16074266985328764,
+      "learning_rate": 3.5496593505559575e-06,
+      "loss": 0.4809,
+      "step": 1808
+    },
+    {
+      "epoch": 1.7838304165639636,
+      "grad_norm": 0.17896578555040946,
+      "learning_rate": 3.5447103035582285e-06,
+      "loss": 0.4621,
+      "step": 1809
+    },
+    {
+      "epoch": 1.784816366773478,
+      "grad_norm": 0.15710714831802694,
+      "learning_rate": 3.53976281381494e-06,
+      "loss": 0.4446,
+      "step": 1810
+    },
+    {
+      "epoch": 1.7858023169829924,
+      "grad_norm": 0.15832639388066821,
+      "learning_rate": 3.5348168866202226e-06,
+      "loss": 0.4598,
+      "step": 1811
+    },
+    {
+      "epoch": 1.7867882671925068,
+      "grad_norm": 0.18073821236033372,
+      "learning_rate": 3.529872527266542e-06,
+      "loss": 0.4631,
+      "step": 1812
+    },
+    {
+      "epoch": 1.7877742174020212,
+      "grad_norm": 0.15627997258945053,
+      "learning_rate": 3.5249297410446836e-06,
+      "loss": 0.4615,
+      "step": 1813
+    },
+    {
+      "epoch": 1.7887601676115357,
+      "grad_norm": 0.15690753303724347,
+      "learning_rate": 3.519988533243749e-06,
+      "loss": 0.4696,
+      "step": 1814
+    },
+    {
+      "epoch": 1.78974611782105,
+      "grad_norm": 0.15726083379428013,
+      "learning_rate": 3.515048909151154e-06,
+      "loss": 0.4839,
+      "step": 1815
+    },
+    {
+      "epoch": 1.7907320680305645,
+      "grad_norm": 0.15438431777406547,
+      "learning_rate": 3.5101108740526134e-06,
+      "loss": 0.4549,
+      "step": 1816
+    },
+    {
+      "epoch": 1.791718018240079,
+      "grad_norm": 0.15512043084625451,
+      "learning_rate": 3.505174433232147e-06,
+      "loss": 0.4491,
+      "step": 1817
+    },
+    {
+      "epoch": 1.7927039684495933,
+      "grad_norm": 0.1633839584996681,
+      "learning_rate": 3.500239591972065e-06,
+      "loss": 0.4543,
+      "step": 1818
+    },
+    {
+      "epoch": 1.7936899186591078,
+      "grad_norm": 0.1616122255325751,
+      "learning_rate": 3.4953063555529703e-06,
+      "loss": 0.486,
+      "step": 1819
+    },
+    {
+      "epoch": 1.7946758688686222,
+      "grad_norm": 0.16712653497862173,
+      "learning_rate": 3.4903747292537467e-06,
+      "loss": 0.4659,
+      "step": 1820
+    },
+    {
+      "epoch": 1.7956618190781366,
+      "grad_norm": 0.16922792213817878,
+      "learning_rate": 3.4854447183515504e-06,
+      "loss": 0.4615,
+      "step": 1821
+    },
+    {
+      "epoch": 1.796647769287651,
+      "grad_norm": 0.16721935928538403,
+      "learning_rate": 3.480516328121817e-06,
+      "loss": 0.4344,
+      "step": 1822
+    },
+    {
+      "epoch": 1.7976337194971654,
+      "grad_norm": 0.15720134451512144,
+      "learning_rate": 3.4755895638382413e-06,
+      "loss": 0.448,
+      "step": 1823
+    },
+    {
+      "epoch": 1.7986196697066799,
+      "grad_norm": 0.1571325895931652,
+      "learning_rate": 3.4706644307727833e-06,
+      "loss": 0.4611,
+      "step": 1824
+    },
+    {
+      "epoch": 1.7996056199161943,
+      "grad_norm": 0.16817426785366024,
+      "learning_rate": 3.465740934195655e-06,
+      "loss": 0.4626,
+      "step": 1825
+    },
+    {
+      "epoch": 1.8005915701257087,
+      "grad_norm": 0.16516151765217899,
+      "learning_rate": 3.460819079375315e-06,
+      "loss": 0.4655,
+      "step": 1826
+    },
+    {
+      "epoch": 1.8015775203352231,
+      "grad_norm": 0.16229957509756535,
+      "learning_rate": 3.4558988715784677e-06,
+      "loss": 0.4675,
+      "step": 1827
+    },
+    {
+      "epoch": 1.8025634705447375,
+      "grad_norm": 0.15799579555997448,
+      "learning_rate": 3.4509803160700562e-06,
+      "loss": 0.4684,
+      "step": 1828
+    },
+    {
+      "epoch": 1.803549420754252,
+      "grad_norm": 0.1662848802427933,
+      "learning_rate": 3.4460634181132534e-06,
+      "loss": 0.4579,
+      "step": 1829
+    },
+    {
+      "epoch": 1.8045353709637664,
+      "grad_norm": 0.16554653297373703,
+      "learning_rate": 3.4411481829694627e-06,
+      "loss": 0.4767,
+      "step": 1830
+    },
+    {
+      "epoch": 1.8055213211732808,
+      "grad_norm": 0.17551491880662762,
+      "learning_rate": 3.4362346158982985e-06,
+      "loss": 0.4516,
+      "step": 1831
+    },
+    {
+      "epoch": 1.8065072713827952,
+      "grad_norm": 0.1599166072934886,
+      "learning_rate": 3.4313227221576008e-06,
+      "loss": 0.4843,
+      "step": 1832
+    },
+    {
+      "epoch": 1.8074932215923096,
+      "grad_norm": 0.15895594620786174,
+      "learning_rate": 3.4264125070034115e-06,
+      "loss": 0.4666,
+      "step": 1833
+    },
+    {
+      "epoch": 1.808479171801824,
+      "grad_norm": 0.16165619505400786,
+      "learning_rate": 3.4215039756899836e-06,
+      "loss": 0.455,
+      "step": 1834
+    },
+    {
+      "epoch": 1.8094651220113385,
+      "grad_norm": 0.15832254009515392,
+      "learning_rate": 3.4165971334697633e-06,
+      "loss": 0.4558,
+      "step": 1835
+    },
+    {
+      "epoch": 1.8104510722208529,
+      "grad_norm": 0.15990852763698515,
+      "learning_rate": 3.411691985593387e-06,
+      "loss": 0.4596,
+      "step": 1836
+    },
+    {
+      "epoch": 1.8114370224303673,
+      "grad_norm": 0.15622449666788027,
+      "learning_rate": 3.406788537309685e-06,
+      "loss": 0.4537,
+      "step": 1837
+    },
+    {
+      "epoch": 1.8124229726398817,
+      "grad_norm": 0.17240945192540147,
+      "learning_rate": 3.401886793865663e-06,
+      "loss": 0.4598,
+      "step": 1838
+    },
+    {
+      "epoch": 1.8134089228493961,
+      "grad_norm": 0.158165938089437,
+      "learning_rate": 3.3969867605065055e-06,
+      "loss": 0.4489,
+      "step": 1839
+    },
+    {
+      "epoch": 1.8143948730589106,
+      "grad_norm": 0.1604750780707797,
+      "learning_rate": 3.3920884424755674e-06,
+      "loss": 0.4818,
+      "step": 1840
+    },
+    {
+      "epoch": 1.815380823268425,
+      "grad_norm": 0.15646417111600883,
+      "learning_rate": 3.3871918450143647e-06,
+      "loss": 0.4604,
+      "step": 1841
+    },
+    {
+      "epoch": 1.8163667734779394,
+      "grad_norm": 0.15902233562082224,
+      "learning_rate": 3.3822969733625747e-06,
+      "loss": 0.431,
+      "step": 1842
+    },
+    {
+      "epoch": 1.8173527236874538,
+      "grad_norm": 0.17308042764576753,
+      "learning_rate": 3.37740383275803e-06,
+      "loss": 0.4647,
+      "step": 1843
+    },
+    {
+      "epoch": 1.8183386738969682,
+      "grad_norm": 0.16678974761648963,
+      "learning_rate": 3.3725124284367074e-06,
+      "loss": 0.461,
+      "step": 1844
+    },
+    {
+      "epoch": 1.8193246241064827,
+      "grad_norm": 0.16282534173302918,
+      "learning_rate": 3.3676227656327277e-06,
+      "loss": 0.4458,
+      "step": 1845
+    },
+    {
+      "epoch": 1.820310574315997,
+      "grad_norm": 0.15062473045003175,
+      "learning_rate": 3.3627348495783445e-06,
+      "loss": 0.4575,
+      "step": 1846
+    },
+    {
+      "epoch": 1.8212965245255115,
+      "grad_norm": 0.18143489354081666,
+      "learning_rate": 3.3578486855039488e-06,
+      "loss": 0.4439,
+      "step": 1847
+    },
+    {
+      "epoch": 1.822282474735026,
+      "grad_norm": 0.15234651761383963,
+      "learning_rate": 3.352964278638051e-06,
+      "loss": 0.4602,
+      "step": 1848
+    },
+    {
+      "epoch": 1.8232684249445403,
+      "grad_norm": 0.15984031597904724,
+      "learning_rate": 3.3480816342072853e-06,
+      "loss": 0.4689,
+      "step": 1849
+    },
+    {
+      "epoch": 1.8242543751540548,
+      "grad_norm": 0.1628495763093985,
+      "learning_rate": 3.343200757436399e-06,
+      "loss": 0.4642,
+      "step": 1850
+    },
+    {
+      "epoch": 1.8252403253635692,
+      "grad_norm": 0.15565905371256994,
+      "learning_rate": 3.338321653548244e-06,
+      "loss": 0.4618,
+      "step": 1851
+    },
+    {
+      "epoch": 1.8262262755730836,
+      "grad_norm": 0.16164990092885384,
+      "learning_rate": 3.3334443277637786e-06,
+      "loss": 0.4638,
+      "step": 1852
+    },
+    {
+      "epoch": 1.827212225782598,
+      "grad_norm": 0.1577594623352456,
+      "learning_rate": 3.3285687853020604e-06,
+      "loss": 0.4724,
+      "step": 1853
+    },
+    {
+      "epoch": 1.8281981759921124,
+      "grad_norm": 0.1625290547710255,
+      "learning_rate": 3.3236950313802334e-06,
+      "loss": 0.4676,
+      "step": 1854
+    },
+    {
+      "epoch": 1.8291841262016268,
+      "grad_norm": 0.15682998914590174,
+      "learning_rate": 3.318823071213534e-06,
+      "loss": 0.4641,
+      "step": 1855
+    },
+    {
+      "epoch": 1.8301700764111413,
+      "grad_norm": 0.16434356883191334,
+      "learning_rate": 3.313952910015274e-06,
+      "loss": 0.4738,
+      "step": 1856
+    },
+    {
+      "epoch": 1.8311560266206557,
+      "grad_norm": 0.16041096783467473,
+      "learning_rate": 3.3090845529968414e-06,
+      "loss": 0.4554,
+      "step": 1857
+    },
+    {
+      "epoch": 1.83214197683017,
+      "grad_norm": 0.15907413737106404,
+      "learning_rate": 3.3042180053676937e-06,
+      "loss": 0.4767,
+      "step": 1858
+    },
+    {
+      "epoch": 1.8331279270396845,
+      "grad_norm": 0.18226908874849415,
+      "learning_rate": 3.2993532723353548e-06,
+      "loss": 0.4808,
+      "step": 1859
+    },
+    {
+      "epoch": 1.834113877249199,
+      "grad_norm": 0.1570027521805479,
+      "learning_rate": 3.2944903591054033e-06,
+      "loss": 0.4534,
+      "step": 1860
+    },
+    {
+      "epoch": 1.8350998274587134,
+      "grad_norm": 0.16304082965163516,
+      "learning_rate": 3.2896292708814736e-06,
+      "loss": 0.4533,
+      "step": 1861
+    },
+    {
+      "epoch": 1.8360857776682278,
+      "grad_norm": 0.15870440809124775,
+      "learning_rate": 3.284770012865245e-06,
+      "loss": 0.4522,
+      "step": 1862
+    },
+    {
+      "epoch": 1.8370717278777422,
+      "grad_norm": 0.15842408817291853,
+      "learning_rate": 3.279912590256438e-06,
+      "loss": 0.459,
+      "step": 1863
+    },
+    {
+      "epoch": 1.8380576780872566,
+      "grad_norm": 0.15994740780885353,
+      "learning_rate": 3.275057008252809e-06,
+      "loss": 0.4503,
+      "step": 1864
+    },
+    {
+      "epoch": 1.839043628296771,
+      "grad_norm": 0.1580316715220139,
+      "learning_rate": 3.270203272050149e-06,
+      "loss": 0.4585,
+      "step": 1865
+    },
+    {
+      "epoch": 1.8400295785062855,
+      "grad_norm": 0.19051325191228838,
+      "learning_rate": 3.265351386842271e-06,
+      "loss": 0.4579,
+      "step": 1866
+    },
+    {
+      "epoch": 1.8410155287157999,
+      "grad_norm": 0.1566686670213712,
+      "learning_rate": 3.2605013578210033e-06,
+      "loss": 0.4541,
+      "step": 1867
+    },
+    {
+      "epoch": 1.8420014789253143,
+      "grad_norm": 0.1560100801294482,
+      "learning_rate": 3.2556531901761945e-06,
+      "loss": 0.4591,
+      "step": 1868
+    },
+    {
+      "epoch": 1.8429874291348287,
+      "grad_norm": 0.1588568819489083,
+      "learning_rate": 3.250806889095698e-06,
+      "loss": 0.4564,
+      "step": 1869
+    },
+    {
+      "epoch": 1.8439733793443431,
+      "grad_norm": 0.157291542776539,
+      "learning_rate": 3.2459624597653703e-06,
+      "loss": 0.4668,
+      "step": 1870
+    },
+    {
+      "epoch": 1.8449593295538576,
+      "grad_norm": 0.1601965240155022,
+      "learning_rate": 3.241119907369068e-06,
+      "loss": 0.4768,
+      "step": 1871
+    },
+    {
+      "epoch": 1.845945279763372,
+      "grad_norm": 0.157013646106603,
+      "learning_rate": 3.2362792370886325e-06,
+      "loss": 0.4704,
+      "step": 1872
+    },
+    {
+      "epoch": 1.8469312299728864,
+      "grad_norm": 0.21012822042122972,
+      "learning_rate": 3.231440454103896e-06,
+      "loss": 0.467,
+      "step": 1873
+    },
+    {
+      "epoch": 1.8479171801824008,
+      "grad_norm": 0.1577460006053354,
+      "learning_rate": 3.226603563592672e-06,
+      "loss": 0.4682,
+      "step": 1874
+    },
+    {
+      "epoch": 1.8489031303919152,
+      "grad_norm": 0.1571004003202859,
+      "learning_rate": 3.2217685707307454e-06,
+      "loss": 0.4615,
+      "step": 1875
+    },
+    {
+      "epoch": 1.8498890806014296,
+      "grad_norm": 0.1684311540483634,
+      "learning_rate": 3.2169354806918773e-06,
+      "loss": 0.4494,
+      "step": 1876
+    },
+    {
+      "epoch": 1.850875030810944,
+      "grad_norm": 0.15444772127631406,
+      "learning_rate": 3.21210429864778e-06,
+      "loss": 0.4428,
+      "step": 1877
+    },
+    {
+      "epoch": 1.8518609810204585,
+      "grad_norm": 0.15973429963787777,
+      "learning_rate": 3.2072750297681375e-06,
+      "loss": 0.4717,
+      "step": 1878
+    },
+    {
+      "epoch": 1.852846931229973,
+      "grad_norm": 0.1525158645228592,
+      "learning_rate": 3.2024476792205783e-06,
+      "loss": 0.4578,
+      "step": 1879
+    },
+    {
+      "epoch": 1.8538328814394873,
+      "grad_norm": 0.1655696267227171,
+      "learning_rate": 3.1976222521706834e-06,
+      "loss": 0.4763,
+      "step": 1880
+    },
+    {
+      "epoch": 1.8548188316490017,
+      "grad_norm": 0.1540107823048566,
+      "learning_rate": 3.1927987537819717e-06,
+      "loss": 0.4506,
+      "step": 1881
+    },
+    {
+      "epoch": 1.8558047818585162,
+      "grad_norm": 0.1521294452532488,
+      "learning_rate": 3.1879771892158972e-06,
+      "loss": 0.4464,
+      "step": 1882
+    },
+    {
+      "epoch": 1.8567907320680306,
+      "grad_norm": 0.16265111807762048,
+      "learning_rate": 3.18315756363185e-06,
+      "loss": 0.4716,
+      "step": 1883
+    },
+    {
+      "epoch": 1.857776682277545,
+      "grad_norm": 0.18365951500532943,
+      "learning_rate": 3.178339882187142e-06,
+      "loss": 0.4762,
+      "step": 1884
+    },
+    {
+      "epoch": 1.8587626324870594,
+      "grad_norm": 0.19219541326891357,
+      "learning_rate": 3.173524150037003e-06,
+      "loss": 0.4549,
+      "step": 1885
+    },
+    {
+      "epoch": 1.8597485826965738,
+      "grad_norm": 0.15866139316926775,
+      "learning_rate": 3.1687103723345824e-06,
+      "loss": 0.4625,
+      "step": 1886
+    },
+    {
+      "epoch": 1.8607345329060883,
+      "grad_norm": 0.16772825255039564,
+      "learning_rate": 3.163898554230932e-06,
+      "loss": 0.4661,
+      "step": 1887
+    },
+    {
+      "epoch": 1.8617204831156027,
+      "grad_norm": 0.1605491227208391,
+      "learning_rate": 3.1590887008750092e-06,
+      "loss": 0.4404,
+      "step": 1888
+    },
+    {
+      "epoch": 1.862706433325117,
+      "grad_norm": 0.1495008561900215,
+      "learning_rate": 3.154280817413672e-06,
+      "loss": 0.4455,
+      "step": 1889
+    },
+    {
+      "epoch": 1.8636923835346315,
+      "grad_norm": 0.15886459063277308,
+      "learning_rate": 3.1494749089916652e-06,
+      "loss": 0.4537,
+      "step": 1890
+    },
+    {
+      "epoch": 1.864678333744146,
+      "grad_norm": 0.1640140958159717,
+      "learning_rate": 3.144670980751625e-06,
+      "loss": 0.4619,
+      "step": 1891
+    },
+    {
+      "epoch": 1.8656642839536604,
+      "grad_norm": 0.15834493781763131,
+      "learning_rate": 3.139869037834064e-06,
+      "loss": 0.468,
+      "step": 1892
+    },
+    {
+      "epoch": 1.8666502341631748,
+      "grad_norm": 0.16134168749488406,
+      "learning_rate": 3.1350690853773746e-06,
+      "loss": 0.4398,
+      "step": 1893
+    },
+    {
+      "epoch": 1.8676361843726892,
+      "grad_norm": 0.2253324483513052,
+      "learning_rate": 3.1302711285178156e-06,
+      "loss": 0.4507,
+      "step": 1894
+    },
+    {
+      "epoch": 1.8686221345822036,
+      "grad_norm": 0.16004373862290389,
+      "learning_rate": 3.125475172389515e-06,
+      "loss": 0.4777,
+      "step": 1895
+    },
+    {
+      "epoch": 1.869608084791718,
+      "grad_norm": 0.15471993032254314,
+      "learning_rate": 3.120681222124457e-06,
+      "loss": 0.451,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8705940350012324,
+      "grad_norm": 0.3721044055142246,
+      "learning_rate": 3.115889282852477e-06,
+      "loss": 0.4748,
+      "step": 1897
+    },
+    {
+      "epoch": 1.8715799852107469,
+      "grad_norm": 0.1592218948443952,
+      "learning_rate": 3.1110993597012616e-06,
+      "loss": 0.4709,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8725659354202613,
+      "grad_norm": 0.15870421761671272,
+      "learning_rate": 3.106311457796341e-06,
+      "loss": 0.4398,
+      "step": 1899
+    },
+    {
+      "epoch": 1.8735518856297757,
+      "grad_norm": 0.15730969558231842,
+      "learning_rate": 3.1015255822610794e-06,
+      "loss": 0.4632,
+      "step": 1900
+    },
+    {
+      "epoch": 1.8745378358392901,
+      "grad_norm": 0.16061188271156304,
+      "learning_rate": 3.0967417382166777e-06,
+      "loss": 0.4739,
+      "step": 1901
+    },
+    {
+      "epoch": 1.8755237860488045,
+      "grad_norm": 0.15793561334312298,
+      "learning_rate": 3.0919599307821556e-06,
+      "loss": 0.4659,
+      "step": 1902
+    },
+    {
+      "epoch": 1.876509736258319,
+      "grad_norm": 0.16235126158620328,
+      "learning_rate": 3.0871801650743583e-06,
+      "loss": 0.4707,
+      "step": 1903
+    },
+    {
+      "epoch": 1.8774956864678334,
+      "grad_norm": 0.16356268873089355,
+      "learning_rate": 3.082402446207946e-06,
+      "loss": 0.4748,
+      "step": 1904
+    },
+    {
+      "epoch": 1.8784816366773478,
+      "grad_norm": 0.15619435962510766,
+      "learning_rate": 3.0776267792953886e-06,
+      "loss": 0.4399,
+      "step": 1905
+    },
+    {
+      "epoch": 1.8794675868868622,
+      "grad_norm": 0.16595363086120873,
+      "learning_rate": 3.07285316944696e-06,
+      "loss": 0.4562,
+      "step": 1906
+    },
+    {
+      "epoch": 1.8804535370963766,
+      "grad_norm": 0.15997234378256922,
+      "learning_rate": 3.0680816217707293e-06,
+      "loss": 0.4632,
+      "step": 1907
+    },
+    {
+      "epoch": 1.881439487305891,
+      "grad_norm": 0.16296615025499692,
+      "learning_rate": 3.063312141372566e-06,
+      "loss": 0.4649,
+      "step": 1908
+    },
+    {
+      "epoch": 1.8824254375154055,
+      "grad_norm": 0.16153964356530928,
+      "learning_rate": 3.0585447333561206e-06,
+      "loss": 0.4659,
+      "step": 1909
+    },
+    {
+      "epoch": 1.88341138772492,
+      "grad_norm": 0.1555280681334106,
+      "learning_rate": 3.0537794028228327e-06,
+      "loss": 0.4662,
+      "step": 1910
+    },
+    {
+      "epoch": 1.8843973379344343,
+      "grad_norm": 0.15917860361428368,
+      "learning_rate": 3.049016154871914e-06,
+      "loss": 0.4555,
+      "step": 1911
+    },
+    {
+      "epoch": 1.8853832881439487,
+      "grad_norm": 0.15686598908876545,
+      "learning_rate": 3.0442549946003475e-06,
+      "loss": 0.4671,
+      "step": 1912
+    },
+    {
+      "epoch": 1.8863692383534632,
+      "grad_norm": 0.15708166799382442,
+      "learning_rate": 3.0394959271028836e-06,
+      "loss": 0.4767,
+      "step": 1913
+    },
+    {
+      "epoch": 1.8873551885629776,
+      "grad_norm": 0.16339088997013207,
+      "learning_rate": 3.0347389574720355e-06,
+      "loss": 0.4668,
+      "step": 1914
+    },
+    {
+      "epoch": 1.888341138772492,
+      "grad_norm": 0.15546193949380208,
+      "learning_rate": 3.029984090798067e-06,
+      "loss": 0.4565,
+      "step": 1915
+    },
+    {
+      "epoch": 1.8893270889820064,
+      "grad_norm": 0.15540520000015712,
+      "learning_rate": 3.025231332168999e-06,
+      "loss": 0.4347,
+      "step": 1916
+    },
+    {
+      "epoch": 1.8903130391915208,
+      "grad_norm": 0.15639257919675925,
+      "learning_rate": 3.0204806866705847e-06,
+      "loss": 0.4656,
+      "step": 1917
+    },
+    {
+      "epoch": 1.8912989894010352,
+      "grad_norm": 0.1590228273368216,
+      "learning_rate": 3.015732159386329e-06,
+      "loss": 0.4743,
+      "step": 1918
+    },
+    {
+      "epoch": 1.8922849396105497,
+      "grad_norm": 0.1581612910456602,
+      "learning_rate": 3.0109857553974598e-06,
+      "loss": 0.4666,
+      "step": 1919
+    },
+    {
+      "epoch": 1.893270889820064,
+      "grad_norm": 0.18049981454409514,
+      "learning_rate": 3.006241479782941e-06,
+      "loss": 0.4556,
+      "step": 1920
+    },
+    {
+      "epoch": 1.8942568400295785,
+      "grad_norm": 0.16070363536732327,
+      "learning_rate": 3.0014993376194555e-06,
+      "loss": 0.457,
+      "step": 1921
+    },
+    {
+      "epoch": 1.895242790239093,
+      "grad_norm": 0.165616397602939,
+      "learning_rate": 2.9967593339814003e-06,
+      "loss": 0.4563,
+      "step": 1922
+    },
+    {
+      "epoch": 1.8962287404486073,
+      "grad_norm": 0.1559989880531137,
+      "learning_rate": 2.992021473940888e-06,
+      "loss": 0.4689,
+      "step": 1923
+    },
+    {
+      "epoch": 1.8972146906581218,
+      "grad_norm": 0.15719253856621765,
+      "learning_rate": 2.9872857625677382e-06,
+      "loss": 0.4743,
+      "step": 1924
+    },
+    {
+      "epoch": 1.8982006408676362,
+      "grad_norm": 0.15945375444114104,
+      "learning_rate": 2.982552204929467e-06,
+      "loss": 0.4548,
+      "step": 1925
+    },
+    {
+      "epoch": 1.8991865910771506,
+      "grad_norm": 0.18053807856357737,
+      "learning_rate": 2.9778208060912915e-06,
+      "loss": 0.4808,
+      "step": 1926
+    },
+    {
+      "epoch": 1.900172541286665,
+      "grad_norm": 0.1695676212701229,
+      "learning_rate": 2.9730915711161125e-06,
+      "loss": 0.4611,
+      "step": 1927
+    },
+    {
+      "epoch": 1.9011584914961794,
+      "grad_norm": 0.1577863548799166,
+      "learning_rate": 2.968364505064518e-06,
+      "loss": 0.4669,
+      "step": 1928
+    },
+    {
+      "epoch": 1.9021444417056939,
+      "grad_norm": 0.16087139519802168,
+      "learning_rate": 2.963639612994779e-06,
+      "loss": 0.4398,
+      "step": 1929
+    },
+    {
+      "epoch": 1.9031303919152083,
+      "grad_norm": 0.15507935944597756,
+      "learning_rate": 2.9589168999628335e-06,
+      "loss": 0.4513,
+      "step": 1930
+    },
+    {
+      "epoch": 1.9041163421247227,
+      "grad_norm": 0.21589745345780148,
+      "learning_rate": 2.9541963710222944e-06,
+      "loss": 0.4655,
+      "step": 1931
+    },
+    {
+      "epoch": 1.9051022923342371,
+      "grad_norm": 0.15433199589637409,
+      "learning_rate": 2.9494780312244293e-06,
+      "loss": 0.451,
+      "step": 1932
+    },
+    {
+      "epoch": 1.9060882425437515,
+      "grad_norm": 0.15924290014028386,
+      "learning_rate": 2.9447618856181713e-06,
+      "loss": 0.4584,
+      "step": 1933
+    },
+    {
+      "epoch": 1.907074192753266,
+      "grad_norm": 0.16140321856670797,
+      "learning_rate": 2.9400479392501015e-06,
+      "loss": 0.4595,
+      "step": 1934
+    },
+    {
+      "epoch": 1.9080601429627804,
+      "grad_norm": 0.16019290065835354,
+      "learning_rate": 2.935336197164449e-06,
+      "loss": 0.4793,
+      "step": 1935
+    },
+    {
+      "epoch": 1.9090460931722948,
+      "grad_norm": 0.16092143984089813,
+      "learning_rate": 2.9306266644030836e-06,
+      "loss": 0.4739,
+      "step": 1936
+    },
+    {
+      "epoch": 1.9100320433818092,
+      "grad_norm": 0.1633958525890816,
+      "learning_rate": 2.925919346005512e-06,
+      "loss": 0.4691,
+      "step": 1937
+    },
+    {
+      "epoch": 1.9110179935913236,
+      "grad_norm": 0.15500558425847583,
+      "learning_rate": 2.9212142470088665e-06,
+      "loss": 0.4546,
+      "step": 1938
+    },
+    {
+      "epoch": 1.912003943800838,
+      "grad_norm": 0.1482116239473967,
+      "learning_rate": 2.916511372447912e-06,
+      "loss": 0.4374,
+      "step": 1939
+    },
+    {
+      "epoch": 1.9129898940103525,
+      "grad_norm": 0.15839043230733438,
+      "learning_rate": 2.9118107273550295e-06,
+      "loss": 0.4529,
+      "step": 1940
+    },
+    {
+      "epoch": 1.9139758442198669,
+      "grad_norm": 0.17484570335708627,
+      "learning_rate": 2.907112316760213e-06,
+      "loss": 0.454,
+      "step": 1941
+    },
+    {
+      "epoch": 1.9149617944293813,
+      "grad_norm": 0.16851150199128412,
+      "learning_rate": 2.9024161456910704e-06,
+      "loss": 0.4565,
+      "step": 1942
+    },
+    {
+      "epoch": 1.9159477446388957,
+      "grad_norm": 0.15688125899118022,
+      "learning_rate": 2.8977222191728015e-06,
+      "loss": 0.4519,
+      "step": 1943
+    },
+    {
+      "epoch": 1.9169336948484101,
+      "grad_norm": 0.15899098416715704,
+      "learning_rate": 2.89303054222822e-06,
+      "loss": 0.4529,
+      "step": 1944
+    },
+    {
+      "epoch": 1.9179196450579246,
+      "grad_norm": 0.1610388113835398,
+      "learning_rate": 2.8883411198777224e-06,
+      "loss": 0.4635,
+      "step": 1945
+    },
+    {
+      "epoch": 1.918905595267439,
+      "grad_norm": 0.16148759270031587,
+      "learning_rate": 2.883653957139294e-06,
+      "loss": 0.4536,
+      "step": 1946
+    },
+    {
+      "epoch": 1.9198915454769534,
+      "grad_norm": 0.1575978985313557,
+      "learning_rate": 2.8789690590285048e-06,
+      "loss": 0.4508,
+      "step": 1947
+    },
+    {
+      "epoch": 1.9208774956864678,
+      "grad_norm": 0.16225728719964383,
+      "learning_rate": 2.8742864305584982e-06,
+      "loss": 0.4568,
+      "step": 1948
+    },
+    {
+      "epoch": 1.9218634458959822,
+      "grad_norm": 0.15894418451652362,
+      "learning_rate": 2.8696060767399926e-06,
+      "loss": 0.4653,
+      "step": 1949
+    },
+    {
+      "epoch": 1.9228493961054967,
+      "grad_norm": 0.15910831562107275,
+      "learning_rate": 2.8649280025812688e-06,
+      "loss": 0.475,
+      "step": 1950
+    },
+    {
+      "epoch": 1.923835346315011,
+      "grad_norm": 0.15411775631030855,
+      "learning_rate": 2.860252213088168e-06,
+      "loss": 0.459,
+      "step": 1951
+    },
+    {
+      "epoch": 1.9248212965245255,
+      "grad_norm": 0.15869065524896728,
+      "learning_rate": 2.8555787132640943e-06,
+      "loss": 0.4604,
+      "step": 1952
+    },
+    {
+      "epoch": 1.92580724673404,
+      "grad_norm": 0.15371324508352271,
+      "learning_rate": 2.8509075081099913e-06,
+      "loss": 0.4487,
+      "step": 1953
+    },
+    {
+      "epoch": 1.9267931969435543,
+      "grad_norm": 0.1663553768569787,
+      "learning_rate": 2.84623860262435e-06,
+      "loss": 0.4455,
+      "step": 1954
+    },
+    {
+      "epoch": 1.9277791471530688,
+      "grad_norm": 0.16223804975948958,
+      "learning_rate": 2.8415720018032066e-06,
+      "loss": 0.4482,
+      "step": 1955
+    },
+    {
+      "epoch": 1.9287650973625832,
+      "grad_norm": 0.16013186345810299,
+      "learning_rate": 2.8369077106401266e-06,
+      "loss": 0.4574,
+      "step": 1956
+    },
+    {
+      "epoch": 1.9297510475720976,
+      "grad_norm": 0.1569134362810151,
+      "learning_rate": 2.8322457341262044e-06,
+      "loss": 0.4684,
+      "step": 1957
+    },
+    {
+      "epoch": 1.930736997781612,
+      "grad_norm": 0.1624895711166642,
+      "learning_rate": 2.827586077250052e-06,
+      "loss": 0.4536,
+      "step": 1958
+    },
+    {
+      "epoch": 1.9317229479911264,
+      "grad_norm": 0.1571745912877177,
+      "learning_rate": 2.8229287449978105e-06,
+      "loss": 0.4618,
+      "step": 1959
+    },
+    {
+      "epoch": 1.9327088982006408,
+      "grad_norm": 0.15917210331188558,
+      "learning_rate": 2.8182737423531264e-06,
+      "loss": 0.4686,
+      "step": 1960
+    },
+    {
+      "epoch": 1.9336948484101553,
+      "grad_norm": 0.1560915049196943,
+      "learning_rate": 2.813621074297155e-06,
+      "loss": 0.4443,
+      "step": 1961
+    },
+    {
+      "epoch": 1.9346807986196697,
+      "grad_norm": 0.15844711297710645,
+      "learning_rate": 2.808970745808551e-06,
+      "loss": 0.4712,
+      "step": 1962
+    },
+    {
+      "epoch": 1.935666748829184,
+      "grad_norm": 0.16425246698893556,
+      "learning_rate": 2.8043227618634703e-06,
+      "loss": 0.4548,
+      "step": 1963
+    },
+    {
+      "epoch": 1.9366526990386985,
+      "grad_norm": 0.17328731147764184,
+      "learning_rate": 2.799677127435556e-06,
+      "loss": 0.4524,
+      "step": 1964
+    },
+    {
+      "epoch": 1.937638649248213,
+      "grad_norm": 0.16153363030874587,
+      "learning_rate": 2.7950338474959395e-06,
+      "loss": 0.4531,
+      "step": 1965
+    },
+    {
+      "epoch": 1.9386245994577274,
+      "grad_norm": 0.15840085787844096,
+      "learning_rate": 2.790392927013228e-06,
+      "loss": 0.4634,
+      "step": 1966
+    },
+    {
+      "epoch": 1.9396105496672418,
+      "grad_norm": 0.156182280688547,
+      "learning_rate": 2.785754370953515e-06,
+      "loss": 0.4659,
+      "step": 1967
+    },
+    {
+      "epoch": 1.9405964998767562,
+      "grad_norm": 0.1577606130078244,
+      "learning_rate": 2.7811181842803504e-06,
+      "loss": 0.4562,
+      "step": 1968
+    },
+    {
+      "epoch": 1.9415824500862706,
+      "grad_norm": 0.16285074167980573,
+      "learning_rate": 2.7764843719547562e-06,
+      "loss": 0.4673,
+      "step": 1969
+    },
+    {
+      "epoch": 1.942568400295785,
+      "grad_norm": 0.16210584044877782,
+      "learning_rate": 2.7718529389352123e-06,
+      "loss": 0.4427,
+      "step": 1970
+    },
+    {
+      "epoch": 1.9435543505052995,
+      "grad_norm": 0.15264033577467623,
+      "learning_rate": 2.7672238901776537e-06,
+      "loss": 0.4622,
+      "step": 1971
+    },
+    {
+      "epoch": 1.9445403007148139,
+      "grad_norm": 0.1553511328644889,
+      "learning_rate": 2.7625972306354652e-06,
+      "loss": 0.4601,
+      "step": 1972
+    },
+    {
+      "epoch": 1.9455262509243283,
+      "grad_norm": 0.15749846113915345,
+      "learning_rate": 2.7579729652594667e-06,
+      "loss": 0.4631,
+      "step": 1973
+    },
+    {
+      "epoch": 1.9465122011338427,
+      "grad_norm": 0.1523349961098614,
+      "learning_rate": 2.7533510989979272e-06,
+      "loss": 0.4386,
+      "step": 1974
+    },
+    {
+      "epoch": 1.9474981513433571,
+      "grad_norm": 0.15286019953152136,
+      "learning_rate": 2.7487316367965435e-06,
+      "loss": 0.4512,
+      "step": 1975
+    },
+    {
+      "epoch": 1.9484841015528716,
+      "grad_norm": 0.16000652795911166,
+      "learning_rate": 2.7441145835984374e-06,
+      "loss": 0.4665,
+      "step": 1976
+    },
+    {
+      "epoch": 1.949470051762386,
+      "grad_norm": 0.17317625216154026,
+      "learning_rate": 2.739499944344157e-06,
+      "loss": 0.4704,
+      "step": 1977
+    },
+    {
+      "epoch": 1.9504560019719004,
+      "grad_norm": 0.15390488605498684,
+      "learning_rate": 2.7348877239716665e-06,
+      "loss": 0.4502,
+      "step": 1978
+    },
+    {
+      "epoch": 1.9514419521814148,
+      "grad_norm": 0.15929470966946352,
+      "learning_rate": 2.730277927416339e-06,
+      "loss": 0.4646,
+      "step": 1979
+    },
+    {
+      "epoch": 1.9524279023909292,
+      "grad_norm": 0.16268777255975028,
+      "learning_rate": 2.725670559610959e-06,
+      "loss": 0.4726,
+      "step": 1980
+    },
+    {
+      "epoch": 1.9534138526004436,
+      "grad_norm": 0.1701458164230504,
+      "learning_rate": 2.721065625485705e-06,
+      "loss": 0.451,
+      "step": 1981
+    },
+    {
+      "epoch": 1.954399802809958,
+      "grad_norm": 0.15856203668413626,
+      "learning_rate": 2.716463129968161e-06,
+      "loss": 0.4578,
+      "step": 1982
+    },
+    {
+      "epoch": 1.9553857530194725,
+      "grad_norm": 0.15688745928162184,
+      "learning_rate": 2.7118630779832918e-06,
+      "loss": 0.4334,
+      "step": 1983
+    },
+    {
+      "epoch": 1.956371703228987,
+      "grad_norm": 0.15049989939886538,
+      "learning_rate": 2.707265474453452e-06,
+      "loss": 0.4439,
+      "step": 1984
+    },
+    {
+      "epoch": 1.9573576534385013,
+      "grad_norm": 0.15598612242573837,
+      "learning_rate": 2.702670324298375e-06,
+      "loss": 0.4749,
+      "step": 1985
+    },
+    {
+      "epoch": 1.9583436036480157,
+      "grad_norm": 0.15765072424731968,
+      "learning_rate": 2.698077632435173e-06,
+      "loss": 0.4553,
+      "step": 1986
+    },
+    {
+      "epoch": 1.9593295538575302,
+      "grad_norm": 0.15897114221475683,
+      "learning_rate": 2.6934874037783245e-06,
+      "loss": 0.471,
+      "step": 1987
+    },
+    {
+      "epoch": 1.9603155040670446,
+      "grad_norm": 0.15597262811001858,
+      "learning_rate": 2.688899643239668e-06,
+      "loss": 0.4539,
+      "step": 1988
+    },
+    {
+      "epoch": 1.961301454276559,
+      "grad_norm": 0.15694708957284073,
+      "learning_rate": 2.6843143557284055e-06,
+      "loss": 0.4427,
+      "step": 1989
+    },
+    {
+      "epoch": 1.9622874044860734,
+      "grad_norm": 0.16147762390646708,
+      "learning_rate": 2.6797315461510965e-06,
+      "loss": 0.4531,
+      "step": 1990
+    },
+    {
+      "epoch": 1.9632733546955878,
+      "grad_norm": 0.16235553988010826,
+      "learning_rate": 2.6751512194116415e-06,
+      "loss": 0.452,
+      "step": 1991
+    },
+    {
+      "epoch": 1.9642593049051023,
+      "grad_norm": 0.15833390799471556,
+      "learning_rate": 2.6705733804112886e-06,
+      "loss": 0.4545,
+      "step": 1992
+    },
+    {
+      "epoch": 1.9652452551146167,
+      "grad_norm": 0.15819603071568658,
+      "learning_rate": 2.6659980340486225e-06,
+      "loss": 0.4551,
+      "step": 1993
+    },
+    {
+      "epoch": 1.966231205324131,
+      "grad_norm": 0.15336847229022132,
+      "learning_rate": 2.66142518521956e-06,
+      "loss": 0.4269,
+      "step": 1994
+    },
+    {
+      "epoch": 1.9672171555336455,
+      "grad_norm": 0.15976227175164648,
+      "learning_rate": 2.656854838817347e-06,
+      "loss": 0.4716,
+      "step": 1995
+    },
+    {
+      "epoch": 1.96820310574316,
+      "grad_norm": 0.1568909109038371,
+      "learning_rate": 2.6522869997325505e-06,
+      "loss": 0.4582,
+      "step": 1996
+    },
+    {
+      "epoch": 1.9691890559526743,
+      "grad_norm": 0.15471315466033486,
+      "learning_rate": 2.6477216728530553e-06,
+      "loss": 0.448,
+      "step": 1997
+    },
+    {
+      "epoch": 1.9701750061621888,
+      "grad_norm": 0.1568241985290219,
+      "learning_rate": 2.643158863064057e-06,
+      "loss": 0.4525,
+      "step": 1998
+    },
+    {
+      "epoch": 1.9711609563717032,
+      "grad_norm": 0.1552060004095341,
+      "learning_rate": 2.638598575248058e-06,
+      "loss": 0.4552,
+      "step": 1999
+    },
+    {
+      "epoch": 1.9721469065812176,
+      "grad_norm": 0.15262270444362303,
+      "learning_rate": 2.6340408142848607e-06,
+      "loss": 0.4451,
+      "step": 2000
+    },
+    {
+      "epoch": 1.973132856790732,
+      "grad_norm": 0.15713750671861487,
+      "learning_rate": 2.6294855850515687e-06,
+      "loss": 0.4597,
+      "step": 2001
+    },
+    {
+      "epoch": 1.9741188070002464,
+      "grad_norm": 0.16171825414027022,
+      "learning_rate": 2.6249328924225737e-06,
+      "loss": 0.4739,
+      "step": 2002
+    },
+    {
+      "epoch": 1.9751047572097609,
+      "grad_norm": 0.1564439191029133,
+      "learning_rate": 2.6203827412695482e-06,
+      "loss": 0.4519,
+      "step": 2003
+    },
+    {
+      "epoch": 1.9760907074192753,
+      "grad_norm": 0.1555304959709648,
+      "learning_rate": 2.6158351364614487e-06,
+      "loss": 0.4401,
+      "step": 2004
+    },
+    {
+      "epoch": 1.9770766576287897,
+      "grad_norm": 0.15741543443360018,
+      "learning_rate": 2.6112900828645116e-06,
+      "loss": 0.4583,
+      "step": 2005
+    },
+    {
+      "epoch": 1.9780626078383041,
+      "grad_norm": 0.1590870159959881,
+      "learning_rate": 2.606747585342238e-06,
+      "loss": 0.4596,
+      "step": 2006
+    },
+    {
+      "epoch": 1.9790485580478185,
+      "grad_norm": 0.16034416018300318,
+      "learning_rate": 2.602207648755395e-06,
+      "loss": 0.4688,
+      "step": 2007
+    },
+    {
+      "epoch": 1.980034508257333,
+      "grad_norm": 0.15786997284929244,
+      "learning_rate": 2.5976702779620077e-06,
+      "loss": 0.4623,
+      "step": 2008
+    },
+    {
+      "epoch": 1.9810204584668474,
+      "grad_norm": 0.15252651356961516,
+      "learning_rate": 2.59313547781736e-06,
+      "loss": 0.4419,
+      "step": 2009
+    },
+    {
+      "epoch": 1.9820064086763618,
+      "grad_norm": 0.16291103856756803,
+      "learning_rate": 2.5886032531739813e-06,
+      "loss": 0.4622,
+      "step": 2010
+    },
+    {
+      "epoch": 1.9829923588858762,
+      "grad_norm": 0.1583943249207326,
+      "learning_rate": 2.584073608881645e-06,
+      "loss": 0.4578,
+      "step": 2011
+    },
+    {
+      "epoch": 1.9839783090953906,
+      "grad_norm": 0.15954374949997202,
+      "learning_rate": 2.579546549787366e-06,
+      "loss": 0.4709,
+      "step": 2012
+    },
+    {
+      "epoch": 1.984964259304905,
+      "grad_norm": 0.17337484989325813,
+      "learning_rate": 2.575022080735391e-06,
+      "loss": 0.4996,
+      "step": 2013
+    },
+    {
+      "epoch": 1.9859502095144195,
+      "grad_norm": 0.15914435262299276,
+      "learning_rate": 2.5705002065671956e-06,
+      "loss": 0.469,
+      "step": 2014
+    },
+    {
+      "epoch": 1.986936159723934,
+      "grad_norm": 0.15636976490103763,
+      "learning_rate": 2.5659809321214784e-06,
+      "loss": 0.4543,
+      "step": 2015
+    },
+    {
+      "epoch": 1.9879221099334483,
+      "grad_norm": 0.18144576680530625,
+      "learning_rate": 2.5614642622341548e-06,
+      "loss": 0.4543,
+      "step": 2016
+    },
+    {
+      "epoch": 1.9889080601429627,
+      "grad_norm": 0.15069657104321396,
+      "learning_rate": 2.5569502017383585e-06,
+      "loss": 0.4472,
+      "step": 2017
+    },
+    {
+      "epoch": 1.9898940103524771,
+      "grad_norm": 0.1581192101077221,
+      "learning_rate": 2.552438755464427e-06,
+      "loss": 0.4503,
+      "step": 2018
+    },
+    {
+      "epoch": 1.9908799605619916,
+      "grad_norm": 0.15794622540757225,
+      "learning_rate": 2.547929928239895e-06,
+      "loss": 0.4648,
+      "step": 2019
+    },
+    {
+      "epoch": 1.991865910771506,
+      "grad_norm": 0.16638480468830721,
+      "learning_rate": 2.543423724889506e-06,
+      "loss": 0.4699,
+      "step": 2020
+    },
+    {
+      "epoch": 1.9928518609810204,
+      "grad_norm": 0.1638747412332113,
+      "learning_rate": 2.538920150235189e-06,
+      "loss": 0.4483,
+      "step": 2021
+    },
+    {
+      "epoch": 1.9938378111905348,
+      "grad_norm": 0.15361835011940492,
+      "learning_rate": 2.5344192090960594e-06,
+      "loss": 0.4588,
+      "step": 2022
+    },
+    {
+      "epoch": 1.9948237614000492,
+      "grad_norm": 0.17243925193115178,
+      "learning_rate": 2.5299209062884185e-06,
+      "loss": 0.4734,
+      "step": 2023
+    },
+    {
+      "epoch": 1.9958097116095637,
+      "grad_norm": 0.15873734963755304,
+      "learning_rate": 2.5254252466257412e-06,
+      "loss": 0.4612,
+      "step": 2024
+    },
+    {
+      "epoch": 1.996795661819078,
+      "grad_norm": 0.1613122922200011,
+      "learning_rate": 2.5209322349186743e-06,
+      "loss": 0.4694,
+      "step": 2025
+    },
+    {
+      "epoch": 1.9977816120285925,
+      "grad_norm": 0.1580632926224322,
+      "learning_rate": 2.5164418759750337e-06,
+      "loss": 0.453,
+      "step": 2026
+    },
+    {
+      "epoch": 1.998767562238107,
+      "grad_norm": 0.15721020580588313,
+      "learning_rate": 2.5119541745997923e-06,
+      "loss": 0.4544,
+      "step": 2027
+    },
+    {
+      "epoch": 1.9997535124476213,
+      "grad_norm": 0.160681699855714,
+      "learning_rate": 2.507469135595087e-06,
+      "loss": 0.4667,
+      "step": 2028
+    },
+    {
+      "epoch": 2.000739462657136,
+      "grad_norm": 0.16414460372966053,
+      "learning_rate": 2.5029867637601955e-06,
+      "loss": 0.4667,
+      "step": 2029
+    },
+    {
+      "epoch": 2.00172541286665,
+      "grad_norm": 0.15005003920043286,
+      "learning_rate": 2.4985070638915485e-06,
+      "loss": 0.4632,
+      "step": 2030
+    },
+    {
+      "epoch": 2.0027113630761644,
+      "grad_norm": 0.15917893988755794,
+      "learning_rate": 2.494030040782714e-06,
+      "loss": 0.4464,
+      "step": 2031
+    },
+    {
+      "epoch": 2.003697313285679,
+      "grad_norm": 0.15354329145987627,
+      "learning_rate": 2.489555699224401e-06,
+      "loss": 0.4544,
+      "step": 2032
+    },
+    {
+      "epoch": 2.0049297510475723,
+      "grad_norm": 0.19316274103980197,
+      "learning_rate": 2.485084044004445e-06,
+      "loss": 0.4115,
+      "step": 2033
+    },
+    {
+      "epoch": 2.0059157012570865,
+      "grad_norm": 0.1926969418616464,
+      "learning_rate": 2.480615079907804e-06,
+      "loss": 0.4139,
+      "step": 2034
+    },
+    {
+      "epoch": 2.006901651466601,
+      "grad_norm": 0.18292987466593616,
+      "learning_rate": 2.476148811716559e-06,
+      "loss": 0.4275,
+      "step": 2035
+    },
+    {
+      "epoch": 2.0078876016761154,
+      "grad_norm": 0.16420682338832338,
+      "learning_rate": 2.471685244209911e-06,
+      "loss": 0.4048,
+      "step": 2036
+    },
+    {
+      "epoch": 2.00887355188563,
+      "grad_norm": 0.1902366011581597,
+      "learning_rate": 2.4672243821641656e-06,
+      "loss": 0.4222,
+      "step": 2037
+    },
+    {
+      "epoch": 2.009859502095144,
+      "grad_norm": 0.2254711219316246,
+      "learning_rate": 2.4627662303527342e-06,
+      "loss": 0.4154,
+      "step": 2038
+    },
+    {
+      "epoch": 2.010845452304659,
+      "grad_norm": 0.20342286022645653,
+      "learning_rate": 2.458310793546129e-06,
+      "loss": 0.419,
+      "step": 2039
+    },
+    {
+      "epoch": 2.011831402514173,
+      "grad_norm": 0.18952275197647123,
+      "learning_rate": 2.4538580765119563e-06,
+      "loss": 0.4155,
+      "step": 2040
+    },
+    {
+      "epoch": 2.0128173527236877,
+      "grad_norm": 0.17381576228322404,
+      "learning_rate": 2.449408084014912e-06,
+      "loss": 0.4122,
+      "step": 2041
+    },
+    {
+      "epoch": 2.013803302933202,
+      "grad_norm": 0.20050359828116268,
+      "learning_rate": 2.4449608208167774e-06,
+      "loss": 0.3976,
+      "step": 2042
+    },
+    {
+      "epoch": 2.0147892531427165,
+      "grad_norm": 0.17915308930601712,
+      "learning_rate": 2.440516291676413e-06,
+      "loss": 0.4207,
+      "step": 2043
+    },
+    {
+      "epoch": 2.0157752033522307,
+      "grad_norm": 0.19314790795510625,
+      "learning_rate": 2.4360745013497526e-06,
+      "loss": 0.4211,
+      "step": 2044
+    },
+    {
+      "epoch": 2.0167611535617453,
+      "grad_norm": 0.4596523922508372,
+      "learning_rate": 2.431635454589801e-06,
+      "loss": 0.418,
+      "step": 2045
+    },
+    {
+      "epoch": 2.0177471037712595,
+      "grad_norm": 0.17206936744944476,
+      "learning_rate": 2.4271991561466254e-06,
+      "loss": 0.4084,
+      "step": 2046
+    },
+    {
+      "epoch": 2.018733053980774,
+      "grad_norm": 0.16986684157695564,
+      "learning_rate": 2.422765610767354e-06,
+      "loss": 0.4104,
+      "step": 2047
+    },
+    {
+      "epoch": 2.0197190041902884,
+      "grad_norm": 0.1745733524015661,
+      "learning_rate": 2.4183348231961707e-06,
+      "loss": 0.4271,
+      "step": 2048
+    },
+    {
+      "epoch": 2.020704954399803,
+      "grad_norm": 0.178360592939682,
+      "learning_rate": 2.4139067981743014e-06,
+      "loss": 0.4167,
+      "step": 2049
+    },
+    {
+      "epoch": 2.021690904609317,
+      "grad_norm": 0.16354451493957817,
+      "learning_rate": 2.4094815404400196e-06,
+      "loss": 0.4219,
+      "step": 2050
+    },
+    {
+      "epoch": 2.022676854818832,
+      "grad_norm": 0.16264836760005474,
+      "learning_rate": 2.4050590547286423e-06,
+      "loss": 0.3999,
+      "step": 2051
+    },
+    {
+      "epoch": 2.023662805028346,
+      "grad_norm": 0.159353154129677,
+      "learning_rate": 2.400639345772515e-06,
+      "loss": 0.4171,
+      "step": 2052
+    },
+    {
+      "epoch": 2.0246487552378607,
+      "grad_norm": 0.1614189992245902,
+      "learning_rate": 2.396222418301013e-06,
+      "loss": 0.4096,
+      "step": 2053
+    },
+    {
+      "epoch": 2.025634705447375,
+      "grad_norm": 0.17812903834233001,
+      "learning_rate": 2.3918082770405347e-06,
+      "loss": 0.3997,
+      "step": 2054
+    },
+    {
+      "epoch": 2.0266206556568895,
+      "grad_norm": 0.16519033443031572,
+      "learning_rate": 2.3873969267144993e-06,
+      "loss": 0.4209,
+      "step": 2055
+    },
+    {
+      "epoch": 2.0276066058664037,
+      "grad_norm": 0.15897351804597687,
+      "learning_rate": 2.382988372043336e-06,
+      "loss": 0.4082,
+      "step": 2056
+    },
+    {
+      "epoch": 2.0285925560759184,
+      "grad_norm": 0.16366118349259243,
+      "learning_rate": 2.378582617744486e-06,
+      "loss": 0.4199,
+      "step": 2057
+    },
+    {
+      "epoch": 2.0295785062854326,
+      "grad_norm": 0.23354903751658837,
+      "learning_rate": 2.3741796685323916e-06,
+      "loss": 0.4159,
+      "step": 2058
+    },
+    {
+      "epoch": 2.030564456494947,
+      "grad_norm": 0.16766034200865684,
+      "learning_rate": 2.369779529118494e-06,
+      "loss": 0.4163,
+      "step": 2059
+    },
+    {
+      "epoch": 2.0315504067044614,
+      "grad_norm": 0.16722153377411042,
+      "learning_rate": 2.365382204211229e-06,
+      "loss": 0.4207,
+      "step": 2060
+    },
+    {
+      "epoch": 2.032536356913976,
+      "grad_norm": 0.15843660439226942,
+      "learning_rate": 2.3609876985160192e-06,
+      "loss": 0.3986,
+      "step": 2061
+    },
+    {
+      "epoch": 2.0335223071234902,
+      "grad_norm": 0.16596114170737103,
+      "learning_rate": 2.3565960167352686e-06,
+      "loss": 0.417,
+      "step": 2062
+    },
+    {
+      "epoch": 2.034508257333005,
+      "grad_norm": 0.15935028620806083,
+      "learning_rate": 2.352207163568368e-06,
+      "loss": 0.3908,
+      "step": 2063
+    },
+    {
+      "epoch": 2.035494207542519,
+      "grad_norm": 0.1604915683673604,
+      "learning_rate": 2.3478211437116694e-06,
+      "loss": 0.435,
+      "step": 2064
+    },
+    {
+      "epoch": 2.0364801577520337,
+      "grad_norm": 0.161152593080013,
+      "learning_rate": 2.3434379618584986e-06,
+      "loss": 0.4209,
+      "step": 2065
+    },
+    {
+      "epoch": 2.037466107961548,
+      "grad_norm": 0.194730826383451,
+      "learning_rate": 2.3390576226991486e-06,
+      "loss": 0.4152,
+      "step": 2066
+    },
+    {
+      "epoch": 2.0384520581710626,
+      "grad_norm": 0.20785591489633165,
+      "learning_rate": 2.334680130920865e-06,
+      "loss": 0.422,
+      "step": 2067
+    },
+    {
+      "epoch": 2.0394380083805768,
+      "grad_norm": 0.15622260033990645,
+      "learning_rate": 2.3303054912078492e-06,
+      "loss": 0.4087,
+      "step": 2068
+    },
+    {
+      "epoch": 2.0404239585900914,
+      "grad_norm": 0.16398917357758674,
+      "learning_rate": 2.3259337082412446e-06,
+      "loss": 0.4249,
+      "step": 2069
+    },
+    {
+      "epoch": 2.0414099087996056,
+      "grad_norm": 0.1588498966871557,
+      "learning_rate": 2.3215647866991485e-06,
+      "loss": 0.4143,
+      "step": 2070
+    },
+    {
+      "epoch": 2.0423958590091202,
+      "grad_norm": 0.15823541866578061,
+      "learning_rate": 2.3171987312565885e-06,
+      "loss": 0.3929,
+      "step": 2071
+    },
+    {
+      "epoch": 2.0433818092186344,
+      "grad_norm": 0.16394423751651,
+      "learning_rate": 2.3128355465855273e-06,
+      "loss": 0.4191,
+      "step": 2072
+    },
+    {
+      "epoch": 2.044367759428149,
+      "grad_norm": 0.16884273307766906,
+      "learning_rate": 2.308475237354856e-06,
+      "loss": 0.4298,
+      "step": 2073
+    },
+    {
+      "epoch": 2.0453537096376633,
+      "grad_norm": 0.1756595085447255,
+      "learning_rate": 2.3041178082303878e-06,
+      "loss": 0.4213,
+      "step": 2074
+    },
+    {
+      "epoch": 2.046339659847178,
+      "grad_norm": 0.16665318581076202,
+      "learning_rate": 2.2997632638748553e-06,
+      "loss": 0.4113,
+      "step": 2075
+    },
+    {
+      "epoch": 2.047325610056692,
+      "grad_norm": 0.1774429295851841,
+      "learning_rate": 2.295411608947903e-06,
+      "loss": 0.4081,
+      "step": 2076
+    },
+    {
+      "epoch": 2.0483115602662068,
+      "grad_norm": 0.15685031607760505,
+      "learning_rate": 2.291062848106083e-06,
+      "loss": 0.4191,
+      "step": 2077
+    },
+    {
+      "epoch": 2.049297510475721,
+      "grad_norm": 0.165764475113141,
+      "learning_rate": 2.286716986002857e-06,
+      "loss": 0.4302,
+      "step": 2078
+    },
+    {
+      "epoch": 2.0502834606852356,
+      "grad_norm": 0.16502553641738632,
+      "learning_rate": 2.2823740272885742e-06,
+      "loss": 0.3987,
+      "step": 2079
+    },
+    {
+      "epoch": 2.05126941089475,
+      "grad_norm": 0.16071862487005942,
+      "learning_rate": 2.278033976610482e-06,
+      "loss": 0.4264,
+      "step": 2080
+    },
+    {
+      "epoch": 2.0522553611042644,
+      "grad_norm": 0.1656052682627577,
+      "learning_rate": 2.2736968386127196e-06,
+      "loss": 0.4003,
+      "step": 2081
+    },
+    {
+      "epoch": 2.0532413113137786,
+      "grad_norm": 0.15904591972000107,
+      "learning_rate": 2.2693626179363056e-06,
+      "loss": 0.4079,
+      "step": 2082
+    },
+    {
+      "epoch": 2.0542272615232933,
+      "grad_norm": 0.16097964204825316,
+      "learning_rate": 2.265031319219138e-06,
+      "loss": 0.4191,
+      "step": 2083
+    },
+    {
+      "epoch": 2.0552132117328075,
+      "grad_norm": 0.1728884836779008,
+      "learning_rate": 2.260702947095983e-06,
+      "loss": 0.41,
+      "step": 2084
+    },
+    {
+      "epoch": 2.056199161942322,
+      "grad_norm": 0.16295443236310594,
+      "learning_rate": 2.2563775061984844e-06,
+      "loss": 0.3917,
+      "step": 2085
+    },
+    {
+      "epoch": 2.0571851121518363,
+      "grad_norm": 0.1671292038118426,
+      "learning_rate": 2.2520550011551435e-06,
+      "loss": 0.412,
+      "step": 2086
+    },
+    {
+      "epoch": 2.058171062361351,
+      "grad_norm": 0.1583333241805396,
+      "learning_rate": 2.2477354365913212e-06,
+      "loss": 0.3907,
+      "step": 2087
+    },
+    {
+      "epoch": 2.059157012570865,
+      "grad_norm": 0.16190713679960353,
+      "learning_rate": 2.2434188171292313e-06,
+      "loss": 0.4133,
+      "step": 2088
+    },
+    {
+      "epoch": 2.06014296278038,
+      "grad_norm": 0.16277119817080535,
+      "learning_rate": 2.239105147387938e-06,
+      "loss": 0.4099,
+      "step": 2089
+    },
+    {
+      "epoch": 2.061128912989894,
+      "grad_norm": 0.16058151549029703,
+      "learning_rate": 2.2347944319833476e-06,
+      "loss": 0.4168,
+      "step": 2090
+    },
+    {
+      "epoch": 2.0621148631994086,
+      "grad_norm": 0.15709005185388775,
+      "learning_rate": 2.2304866755282044e-06,
+      "loss": 0.396,
+      "step": 2091
+    },
+    {
+      "epoch": 2.063100813408923,
+      "grad_norm": 0.1662005825518739,
+      "learning_rate": 2.226181882632087e-06,
+      "loss": 0.4299,
+      "step": 2092
+    },
+    {
+      "epoch": 2.0640867636184375,
+      "grad_norm": 0.17124812316783214,
+      "learning_rate": 2.2218800579014076e-06,
+      "loss": 0.4178,
+      "step": 2093
+    },
+    {
+      "epoch": 2.0650727138279517,
+      "grad_norm": 0.15962448482388233,
+      "learning_rate": 2.2175812059393926e-06,
+      "loss": 0.4122,
+      "step": 2094
+    },
+    {
+      "epoch": 2.0660586640374663,
+      "grad_norm": 0.16387594165192024,
+      "learning_rate": 2.213285331346095e-06,
+      "loss": 0.422,
+      "step": 2095
+    },
+    {
+      "epoch": 2.0670446142469805,
+      "grad_norm": 0.16554271394019585,
+      "learning_rate": 2.2089924387183774e-06,
+      "loss": 0.4166,
+      "step": 2096
+    },
+    {
+      "epoch": 2.068030564456495,
+      "grad_norm": 0.1713818526717728,
+      "learning_rate": 2.204702532649917e-06,
+      "loss": 0.4227,
+      "step": 2097
+    },
+    {
+      "epoch": 2.0690165146660093,
+      "grad_norm": 0.1621404786261226,
+      "learning_rate": 2.200415617731192e-06,
+      "loss": 0.4115,
+      "step": 2098
+    },
+    {
+      "epoch": 2.070002464875524,
+      "grad_norm": 0.17156265722786548,
+      "learning_rate": 2.1961316985494737e-06,
+      "loss": 0.4127,
+      "step": 2099
+    },
+    {
+      "epoch": 2.070988415085038,
+      "grad_norm": 0.16477435588711076,
+      "learning_rate": 2.19185077968884e-06,
+      "loss": 0.4133,
+      "step": 2100
+    },
+    {
+      "epoch": 2.071974365294553,
+      "grad_norm": 0.15846960416926448,
+      "learning_rate": 2.1875728657301493e-06,
+      "loss": 0.4071,
+      "step": 2101
+    },
+    {
+      "epoch": 2.072960315504067,
+      "grad_norm": 0.17152652490165182,
+      "learning_rate": 2.1832979612510475e-06,
+      "loss": 0.4035,
+      "step": 2102
+    },
+    {
+      "epoch": 2.0739462657135816,
+      "grad_norm": 0.16031949415577315,
+      "learning_rate": 2.17902607082596e-06,
+      "loss": 0.4146,
+      "step": 2103
+    },
+    {
+      "epoch": 2.074932215923096,
+      "grad_norm": 0.16347258382061014,
+      "learning_rate": 2.1747571990260867e-06,
+      "loss": 0.4034,
+      "step": 2104
+    },
+    {
+      "epoch": 2.0759181661326105,
+      "grad_norm": 0.16202541571213971,
+      "learning_rate": 2.170491350419398e-06,
+      "loss": 0.4035,
+      "step": 2105
+    },
+    {
+      "epoch": 2.0769041163421247,
+      "grad_norm": 0.16057032837909369,
+      "learning_rate": 2.166228529570628e-06,
+      "loss": 0.4221,
+      "step": 2106
+    },
+    {
+      "epoch": 2.0778900665516393,
+      "grad_norm": 0.17546009537245488,
+      "learning_rate": 2.1619687410412728e-06,
+      "loss": 0.4091,
+      "step": 2107
+    },
+    {
+      "epoch": 2.0788760167611535,
+      "grad_norm": 0.16193591789438427,
+      "learning_rate": 2.157711989389579e-06,
+      "loss": 0.4215,
+      "step": 2108
+    },
+    {
+      "epoch": 2.079861966970668,
+      "grad_norm": 0.1735122505713575,
+      "learning_rate": 2.1534582791705545e-06,
+      "loss": 0.415,
+      "step": 2109
+    },
+    {
+      "epoch": 2.0808479171801824,
+      "grad_norm": 0.16593601833381807,
+      "learning_rate": 2.149207614935939e-06,
+      "loss": 0.4087,
+      "step": 2110
+    },
+    {
+      "epoch": 2.081833867389697,
+      "grad_norm": 0.16634238973258567,
+      "learning_rate": 2.1449600012342193e-06,
+      "loss": 0.4243,
+      "step": 2111
+    },
+    {
+      "epoch": 2.082819817599211,
+      "grad_norm": 0.16930759969879747,
+      "learning_rate": 2.1407154426106214e-06,
+      "loss": 0.4238,
+      "step": 2112
+    },
+    {
+      "epoch": 2.083805767808726,
+      "grad_norm": 0.16504342842028494,
+      "learning_rate": 2.136473943607097e-06,
+      "loss": 0.3924,
+      "step": 2113
+    },
+    {
+      "epoch": 2.08479171801824,
+      "grad_norm": 0.16081302287822755,
+      "learning_rate": 2.1322355087623264e-06,
+      "loss": 0.413,
+      "step": 2114
+    },
+    {
+      "epoch": 2.0857776682277547,
+      "grad_norm": 0.16661725880499173,
+      "learning_rate": 2.1280001426117053e-06,
+      "loss": 0.42,
+      "step": 2115
+    },
+    {
+      "epoch": 2.086763618437269,
+      "grad_norm": 0.1641100802727998,
+      "learning_rate": 2.1237678496873554e-06,
+      "loss": 0.4155,
+      "step": 2116
+    },
+    {
+      "epoch": 2.0877495686467835,
+      "grad_norm": 0.16466429540472657,
+      "learning_rate": 2.1195386345181033e-06,
+      "loss": 0.4211,
+      "step": 2117
+    },
+    {
+      "epoch": 2.0887355188562977,
+      "grad_norm": 0.16556551109315398,
+      "learning_rate": 2.1153125016294838e-06,
+      "loss": 0.4179,
+      "step": 2118
+    },
+    {
+      "epoch": 2.0897214690658124,
+      "grad_norm": 0.17060075762383226,
+      "learning_rate": 2.1110894555437333e-06,
+      "loss": 0.4223,
+      "step": 2119
+    },
+    {
+      "epoch": 2.0907074192753266,
+      "grad_norm": 0.16006232694808414,
+      "learning_rate": 2.1068695007797853e-06,
+      "loss": 0.4002,
+      "step": 2120
+    },
+    {
+      "epoch": 2.091693369484841,
+      "grad_norm": 0.16500201659889524,
+      "learning_rate": 2.102652641853265e-06,
+      "loss": 0.4227,
+      "step": 2121
+    },
+    {
+      "epoch": 2.0926793196943554,
+      "grad_norm": 0.1651185828677862,
+      "learning_rate": 2.0984388832764853e-06,
+      "loss": 0.4327,
+      "step": 2122
+    },
+    {
+      "epoch": 2.09366526990387,
+      "grad_norm": 0.16420757197401067,
+      "learning_rate": 2.09422822955844e-06,
+      "loss": 0.4138,
+      "step": 2123
+    },
+    {
+      "epoch": 2.0946512201133842,
+      "grad_norm": 0.1791171221018479,
+      "learning_rate": 2.0900206852048065e-06,
+      "loss": 0.4151,
+      "step": 2124
+    },
+    {
+      "epoch": 2.095637170322899,
+      "grad_norm": 0.1624292350658065,
+      "learning_rate": 2.085816254717926e-06,
+      "loss": 0.4221,
+      "step": 2125
+    },
+    {
+      "epoch": 2.096623120532413,
+      "grad_norm": 0.16270660939755127,
+      "learning_rate": 2.0816149425968113e-06,
+      "loss": 0.4024,
+      "step": 2126
+    },
+    {
+      "epoch": 2.0976090707419277,
+      "grad_norm": 0.1590699419416762,
+      "learning_rate": 2.077416753337143e-06,
+      "loss": 0.4077,
+      "step": 2127
+    },
+    {
+      "epoch": 2.098595020951442,
+      "grad_norm": 0.17500208343336374,
+      "learning_rate": 2.073221691431254e-06,
+      "loss": 0.4164,
+      "step": 2128
+    },
+    {
+      "epoch": 2.0995809711609565,
+      "grad_norm": 0.15798359813998206,
+      "learning_rate": 2.0690297613681348e-06,
+      "loss": 0.4107,
+      "step": 2129
+    },
+    {
+      "epoch": 2.1005669213704707,
+      "grad_norm": 0.16072774202565737,
+      "learning_rate": 2.0648409676334173e-06,
+      "loss": 0.4155,
+      "step": 2130
+    },
+    {
+      "epoch": 2.1015528715799854,
+      "grad_norm": 0.16492565635865292,
+      "learning_rate": 2.0606553147093883e-06,
+      "loss": 0.4143,
+      "step": 2131
+    },
+    {
+      "epoch": 2.1025388217894996,
+      "grad_norm": 0.16183912142588155,
+      "learning_rate": 2.0564728070749657e-06,
+      "loss": 0.414,
+      "step": 2132
+    },
+    {
+      "epoch": 2.103524771999014,
+      "grad_norm": 0.1676213094264793,
+      "learning_rate": 2.0522934492057046e-06,
+      "loss": 0.3956,
+      "step": 2133
+    },
+    {
+      "epoch": 2.1045107222085284,
+      "grad_norm": 0.16152895143010376,
+      "learning_rate": 2.0481172455737896e-06,
+      "loss": 0.419,
+      "step": 2134
+    },
+    {
+      "epoch": 2.105496672418043,
+      "grad_norm": 0.16918238818197492,
+      "learning_rate": 2.0439442006480288e-06,
+      "loss": 0.4214,
+      "step": 2135
+    },
+    {
+      "epoch": 2.1064826226275573,
+      "grad_norm": 0.1670519009757173,
+      "learning_rate": 2.039774318893852e-06,
+      "loss": 0.406,
+      "step": 2136
+    },
+    {
+      "epoch": 2.107468572837072,
+      "grad_norm": 0.1632550277509182,
+      "learning_rate": 2.0356076047733026e-06,
+      "loss": 0.4087,
+      "step": 2137
+    },
+    {
+      "epoch": 2.108454523046586,
+      "grad_norm": 0.16539817074272825,
+      "learning_rate": 2.0314440627450333e-06,
+      "loss": 0.4147,
+      "step": 2138
+    },
+    {
+      "epoch": 2.1094404732561007,
+      "grad_norm": 0.16232307861452594,
+      "learning_rate": 2.027283697264311e-06,
+      "loss": 0.4224,
+      "step": 2139
+    },
+    {
+      "epoch": 2.110426423465615,
+      "grad_norm": 0.15828565746283588,
+      "learning_rate": 2.02312651278299e-06,
+      "loss": 0.4023,
+      "step": 2140
+    },
+    {
+      "epoch": 2.1114123736751296,
+      "grad_norm": 0.16119854000248965,
+      "learning_rate": 2.01897251374953e-06,
+      "loss": 0.4095,
+      "step": 2141
+    },
+    {
+      "epoch": 2.1123983238846438,
+      "grad_norm": 0.15787852221084392,
+      "learning_rate": 2.014821704608977e-06,
+      "loss": 0.3925,
+      "step": 2142
+    },
+    {
+      "epoch": 2.1133842740941584,
+      "grad_norm": 0.18047411849302172,
+      "learning_rate": 2.0106740898029707e-06,
+      "loss": 0.4135,
+      "step": 2143
+    },
+    {
+      "epoch": 2.1143702243036726,
+      "grad_norm": 0.16464118652101478,
+      "learning_rate": 2.0065296737697286e-06,
+      "loss": 0.4184,
+      "step": 2144
+    },
+    {
+      "epoch": 2.1153561745131872,
+      "grad_norm": 0.15975791419504334,
+      "learning_rate": 2.0023884609440387e-06,
+      "loss": 0.3998,
+      "step": 2145
+    },
+    {
+      "epoch": 2.1163421247227014,
+      "grad_norm": 0.15825991418701266,
+      "learning_rate": 1.998250455757273e-06,
+      "loss": 0.4033,
+      "step": 2146
+    },
+    {
+      "epoch": 2.117328074932216,
+      "grad_norm": 0.15782928807968785,
+      "learning_rate": 1.994115662637364e-06,
+      "loss": 0.4108,
+      "step": 2147
+    },
+    {
+      "epoch": 2.1183140251417303,
+      "grad_norm": 0.15913137118988563,
+      "learning_rate": 1.9899840860088075e-06,
+      "loss": 0.4152,
+      "step": 2148
+    },
+    {
+      "epoch": 2.119299975351245,
+      "grad_norm": 0.18908050771039311,
+      "learning_rate": 1.9858557302926605e-06,
+      "loss": 0.4168,
+      "step": 2149
+    },
+    {
+      "epoch": 2.120285925560759,
+      "grad_norm": 0.16220131083638675,
+      "learning_rate": 1.9817305999065312e-06,
+      "loss": 0.4192,
+      "step": 2150
+    },
+    {
+      "epoch": 2.1212718757702738,
+      "grad_norm": 0.15743426841699798,
+      "learning_rate": 1.9776086992645765e-06,
+      "loss": 0.4094,
+      "step": 2151
+    },
+    {
+      "epoch": 2.122257825979788,
+      "grad_norm": 0.15562051150111653,
+      "learning_rate": 1.9734900327774976e-06,
+      "loss": 0.4194,
+      "step": 2152
+    },
+    {
+      "epoch": 2.1232437761893026,
+      "grad_norm": 0.1671001502186119,
+      "learning_rate": 1.969374604852535e-06,
+      "loss": 0.4208,
+      "step": 2153
+    },
+    {
+      "epoch": 2.124229726398817,
+      "grad_norm": 0.15953800531347767,
+      "learning_rate": 1.9652624198934637e-06,
+      "loss": 0.3975,
+      "step": 2154
+    },
+    {
+      "epoch": 2.1252156766083314,
+      "grad_norm": 0.1622468362221989,
+      "learning_rate": 1.961153482300589e-06,
+      "loss": 0.4181,
+      "step": 2155
+    },
+    {
+      "epoch": 2.1262016268178456,
+      "grad_norm": 0.16332419993400998,
+      "learning_rate": 1.95704779647074e-06,
+      "loss": 0.4165,
+      "step": 2156
+    },
+    {
+      "epoch": 2.1271875770273603,
+      "grad_norm": 0.15946977213145044,
+      "learning_rate": 1.9529453667972664e-06,
+      "loss": 0.4026,
+      "step": 2157
+    },
+    {
+      "epoch": 2.1281735272368745,
+      "grad_norm": 0.1592368242342468,
+      "learning_rate": 1.948846197670036e-06,
+      "loss": 0.4193,
+      "step": 2158
+    },
+    {
+      "epoch": 2.129159477446389,
+      "grad_norm": 0.1593730327730932,
+      "learning_rate": 1.944750293475428e-06,
+      "loss": 0.4089,
+      "step": 2159
+    },
+    {
+      "epoch": 2.1301454276559033,
+      "grad_norm": 0.15500326421754593,
+      "learning_rate": 1.940657658596321e-06,
+      "loss": 0.3929,
+      "step": 2160
+    },
+    {
+      "epoch": 2.131131377865418,
+      "grad_norm": 0.16470379968502602,
+      "learning_rate": 1.9365682974120996e-06,
+      "loss": 0.404,
+      "step": 2161
+    },
+    {
+      "epoch": 2.132117328074932,
+      "grad_norm": 0.19538808324042145,
+      "learning_rate": 1.9324822142986505e-06,
+      "loss": 0.4246,
+      "step": 2162
+    },
+    {
+      "epoch": 2.133103278284447,
+      "grad_norm": 0.16156242091510378,
+      "learning_rate": 1.928399413628345e-06,
+      "loss": 0.417,
+      "step": 2163
+    },
+    {
+      "epoch": 2.134089228493961,
+      "grad_norm": 0.17127984814270825,
+      "learning_rate": 1.924319899770045e-06,
+      "loss": 0.4109,
+      "step": 2164
+    },
+    {
+      "epoch": 2.1350751787034756,
+      "grad_norm": 0.16546930929036938,
+      "learning_rate": 1.9202436770890958e-06,
+      "loss": 0.3997,
+      "step": 2165
+    },
+    {
+      "epoch": 2.13606112891299,
+      "grad_norm": 0.16028629302094813,
+      "learning_rate": 1.9161707499473196e-06,
+      "loss": 0.4137,
+      "step": 2166
+    },
+    {
+      "epoch": 2.1370470791225045,
+      "grad_norm": 0.16430777544125744,
+      "learning_rate": 1.9121011227030127e-06,
+      "loss": 0.404,
+      "step": 2167
+    },
+    {
+      "epoch": 2.1380330293320187,
+      "grad_norm": 0.16082152748676395,
+      "learning_rate": 1.908034799710941e-06,
+      "loss": 0.4057,
+      "step": 2168
+    },
+    {
+      "epoch": 2.1390189795415333,
+      "grad_norm": 0.16663143353116502,
+      "learning_rate": 1.9039717853223343e-06,
+      "loss": 0.4087,
+      "step": 2169
+    },
+    {
+      "epoch": 2.1400049297510475,
+      "grad_norm": 0.16418114703833722,
+      "learning_rate": 1.8999120838848806e-06,
+      "loss": 0.404,
+      "step": 2170
+    },
+    {
+      "epoch": 2.140990879960562,
+      "grad_norm": 0.1783395541806536,
+      "learning_rate": 1.8958556997427247e-06,
+      "loss": 0.4117,
+      "step": 2171
+    },
+    {
+      "epoch": 2.1419768301700763,
+      "grad_norm": 0.16686800739942645,
+      "learning_rate": 1.891802637236459e-06,
+      "loss": 0.4068,
+      "step": 2172
+    },
+    {
+      "epoch": 2.142962780379591,
+      "grad_norm": 0.16548420893496701,
+      "learning_rate": 1.887752900703127e-06,
+      "loss": 0.419,
+      "step": 2173
+    },
+    {
+      "epoch": 2.143948730589105,
+      "grad_norm": 0.17374274058848405,
+      "learning_rate": 1.8837064944762097e-06,
+      "loss": 0.4032,
+      "step": 2174
+    },
+    {
+      "epoch": 2.14493468079862,
+      "grad_norm": 0.16498670560317108,
+      "learning_rate": 1.8796634228856209e-06,
+      "loss": 0.4044,
+      "step": 2175
+    },
+    {
+      "epoch": 2.145920631008134,
+      "grad_norm": 0.1654843853823051,
+      "learning_rate": 1.8756236902577096e-06,
+      "loss": 0.4319,
+      "step": 2176
+    },
+    {
+      "epoch": 2.1469065812176487,
+      "grad_norm": 0.16025865811828457,
+      "learning_rate": 1.8715873009152558e-06,
+      "loss": 0.4054,
+      "step": 2177
+    },
+    {
+      "epoch": 2.147892531427163,
+      "grad_norm": 0.16962396410394456,
+      "learning_rate": 1.8675542591774554e-06,
+      "loss": 0.4328,
+      "step": 2178
+    },
+    {
+      "epoch": 2.1488784816366775,
+      "grad_norm": 0.16501085841999438,
+      "learning_rate": 1.8635245693599275e-06,
+      "loss": 0.4067,
+      "step": 2179
+    },
+    {
+      "epoch": 2.1498644318461917,
+      "grad_norm": 0.16200967595281812,
+      "learning_rate": 1.8594982357746965e-06,
+      "loss": 0.4038,
+      "step": 2180
+    },
+    {
+      "epoch": 2.1508503820557063,
+      "grad_norm": 0.15857804828382674,
+      "learning_rate": 1.8554752627302052e-06,
+      "loss": 0.401,
+      "step": 2181
+    },
+    {
+      "epoch": 2.1518363322652205,
+      "grad_norm": 0.16178670261990907,
+      "learning_rate": 1.8514556545312945e-06,
+      "loss": 0.4118,
+      "step": 2182
+    },
+    {
+      "epoch": 2.152822282474735,
+      "grad_norm": 0.16406883339538325,
+      "learning_rate": 1.847439415479207e-06,
+      "loss": 0.4154,
+      "step": 2183
+    },
+    {
+      "epoch": 2.1538082326842494,
+      "grad_norm": 0.16252922142205084,
+      "learning_rate": 1.8434265498715758e-06,
+      "loss": 0.4026,
+      "step": 2184
+    },
+    {
+      "epoch": 2.154794182893764,
+      "grad_norm": 0.16317163382582592,
+      "learning_rate": 1.8394170620024337e-06,
+      "loss": 0.4224,
+      "step": 2185
+    },
+    {
+      "epoch": 2.155780133103278,
+      "grad_norm": 0.16038520344326254,
+      "learning_rate": 1.835410956162188e-06,
+      "loss": 0.4266,
+      "step": 2186
+    },
+    {
+      "epoch": 2.156766083312793,
+      "grad_norm": 0.15798096004987605,
+      "learning_rate": 1.8314082366376335e-06,
+      "loss": 0.4039,
+      "step": 2187
+    },
+    {
+      "epoch": 2.157752033522307,
+      "grad_norm": 0.1592308865545846,
+      "learning_rate": 1.8274089077119378e-06,
+      "loss": 0.4068,
+      "step": 2188
+    },
+    {
+      "epoch": 2.1587379837318217,
+      "grad_norm": 0.17288103981264286,
+      "learning_rate": 1.8234129736646461e-06,
+      "loss": 0.4019,
+      "step": 2189
+    },
+    {
+      "epoch": 2.159723933941336,
+      "grad_norm": 0.1609061792043482,
+      "learning_rate": 1.8194204387716675e-06,
+      "loss": 0.4181,
+      "step": 2190
+    },
+    {
+      "epoch": 2.1607098841508505,
+      "grad_norm": 0.16971146139080798,
+      "learning_rate": 1.8154313073052681e-06,
+      "loss": 0.4265,
+      "step": 2191
+    },
+    {
+      "epoch": 2.1616958343603647,
+      "grad_norm": 0.1620419097268328,
+      "learning_rate": 1.8114455835340827e-06,
+      "loss": 0.423,
+      "step": 2192
+    },
+    {
+      "epoch": 2.1626817845698794,
+      "grad_norm": 0.16480446858563771,
+      "learning_rate": 1.8074632717230927e-06,
+      "loss": 0.4185,
+      "step": 2193
+    },
+    {
+      "epoch": 2.1636677347793936,
+      "grad_norm": 0.1681441180898963,
+      "learning_rate": 1.80348437613363e-06,
+      "loss": 0.4136,
+      "step": 2194
+    },
+    {
+      "epoch": 2.164653684988908,
+      "grad_norm": 0.1591729040367542,
+      "learning_rate": 1.7995089010233718e-06,
+      "loss": 0.4102,
+      "step": 2195
+    },
+    {
+      "epoch": 2.1656396351984224,
+      "grad_norm": 0.16875051880378447,
+      "learning_rate": 1.7955368506463338e-06,
+      "loss": 0.3945,
+      "step": 2196
+    },
+    {
+      "epoch": 2.166625585407937,
+      "grad_norm": 0.16056137879840315,
+      "learning_rate": 1.7915682292528685e-06,
+      "loss": 0.4148,
+      "step": 2197
+    },
+    {
+      "epoch": 2.1676115356174512,
+      "grad_norm": 0.16167005788700176,
+      "learning_rate": 1.7876030410896578e-06,
+      "loss": 0.4118,
+      "step": 2198
+    },
+    {
+      "epoch": 2.168597485826966,
+      "grad_norm": 0.16263157826789873,
+      "learning_rate": 1.7836412903997085e-06,
+      "loss": 0.4297,
+      "step": 2199
+    },
+    {
+      "epoch": 2.16958343603648,
+      "grad_norm": 0.15948787430453046,
+      "learning_rate": 1.7796829814223565e-06,
+      "loss": 0.4177,
+      "step": 2200
+    },
+    {
+      "epoch": 2.1705693862459947,
+      "grad_norm": 0.16878216756886477,
+      "learning_rate": 1.7757281183932445e-06,
+      "loss": 0.4158,
+      "step": 2201
+    },
+    {
+      "epoch": 2.171555336455509,
+      "grad_norm": 0.18730391663546847,
+      "learning_rate": 1.771776705544334e-06,
+      "loss": 0.4062,
+      "step": 2202
+    },
+    {
+      "epoch": 2.1725412866650236,
+      "grad_norm": 0.16302883810094165,
+      "learning_rate": 1.7678287471038914e-06,
+      "loss": 0.4336,
+      "step": 2203
+    },
+    {
+      "epoch": 2.1735272368745377,
+      "grad_norm": 0.17081710177280024,
+      "learning_rate": 1.7638842472964923e-06,
+      "loss": 0.4158,
+      "step": 2204
+    },
+    {
+      "epoch": 2.1745131870840524,
+      "grad_norm": 0.1660414253808993,
+      "learning_rate": 1.759943210343007e-06,
+      "loss": 0.4302,
+      "step": 2205
+    },
+    {
+      "epoch": 2.1754991372935666,
+      "grad_norm": 0.16987499571555054,
+      "learning_rate": 1.756005640460598e-06,
+      "loss": 0.4172,
+      "step": 2206
+    },
+    {
+      "epoch": 2.1764850875030812,
+      "grad_norm": 0.16993580402782144,
+      "learning_rate": 1.7520715418627203e-06,
+      "loss": 0.4187,
+      "step": 2207
+    },
+    {
+      "epoch": 2.1774710377125954,
+      "grad_norm": 0.1642323672574239,
+      "learning_rate": 1.7481409187591186e-06,
+      "loss": 0.4219,
+      "step": 2208
+    },
+    {
+      "epoch": 2.17845698792211,
+      "grad_norm": 0.18503526323285635,
+      "learning_rate": 1.7442137753558126e-06,
+      "loss": 0.4249,
+      "step": 2209
+    },
+    {
+      "epoch": 2.1794429381316243,
+      "grad_norm": 0.16407691332661228,
+      "learning_rate": 1.7402901158551006e-06,
+      "loss": 0.4281,
+      "step": 2210
+    },
+    {
+      "epoch": 2.180428888341139,
+      "grad_norm": 0.15999767519315217,
+      "learning_rate": 1.7363699444555532e-06,
+      "loss": 0.4025,
+      "step": 2211
+    },
+    {
+      "epoch": 2.181414838550653,
+      "grad_norm": 0.1649245050856335,
+      "learning_rate": 1.7324532653520082e-06,
+      "loss": 0.4171,
+      "step": 2212
+    },
+    {
+      "epoch": 2.1824007887601677,
+      "grad_norm": 0.16004150884273285,
+      "learning_rate": 1.7285400827355663e-06,
+      "loss": 0.4185,
+      "step": 2213
+    },
+    {
+      "epoch": 2.183386738969682,
+      "grad_norm": 0.16081158077390012,
+      "learning_rate": 1.7246304007935872e-06,
+      "loss": 0.4119,
+      "step": 2214
+    },
+    {
+      "epoch": 2.1843726891791966,
+      "grad_norm": 0.1810009258055357,
+      "learning_rate": 1.7207242237096844e-06,
+      "loss": 0.4156,
+      "step": 2215
+    },
+    {
+      "epoch": 2.1853586393887108,
+      "grad_norm": 0.16472612332180087,
+      "learning_rate": 1.7168215556637208e-06,
+      "loss": 0.4236,
+      "step": 2216
+    },
+    {
+      "epoch": 2.1863445895982254,
+      "grad_norm": 0.16022270898964686,
+      "learning_rate": 1.7129224008318047e-06,
+      "loss": 0.422,
+      "step": 2217
+    },
+    {
+      "epoch": 2.1873305398077396,
+      "grad_norm": 0.18361054917211825,
+      "learning_rate": 1.7090267633862822e-06,
+      "loss": 0.4266,
+      "step": 2218
+    },
+    {
+      "epoch": 2.1883164900172543,
+      "grad_norm": 0.1588835134399213,
+      "learning_rate": 1.7051346474957432e-06,
+      "loss": 0.3957,
+      "step": 2219
+    },
+    {
+      "epoch": 2.1893024402267685,
+      "grad_norm": 0.15886974731578227,
+      "learning_rate": 1.7012460573250034e-06,
+      "loss": 0.4305,
+      "step": 2220
+    },
+    {
+      "epoch": 2.190288390436283,
+      "grad_norm": 0.16631333224059686,
+      "learning_rate": 1.6973609970351029e-06,
+      "loss": 0.4093,
+      "step": 2221
+    },
+    {
+      "epoch": 2.1912743406457973,
+      "grad_norm": 0.17509958145294982,
+      "learning_rate": 1.6934794707833096e-06,
+      "loss": 0.4082,
+      "step": 2222
+    },
+    {
+      "epoch": 2.192260290855312,
+      "grad_norm": 0.1654129724979681,
+      "learning_rate": 1.6896014827231111e-06,
+      "loss": 0.4135,
+      "step": 2223
+    },
+    {
+      "epoch": 2.193246241064826,
+      "grad_norm": 1.0116963699072126,
+      "learning_rate": 1.6857270370042044e-06,
+      "loss": 0.4142,
+      "step": 2224
+    },
+    {
+      "epoch": 2.1942321912743408,
+      "grad_norm": 0.1593508623469355,
+      "learning_rate": 1.6818561377725002e-06,
+      "loss": 0.4094,
+      "step": 2225
+    },
+    {
+      "epoch": 2.195218141483855,
+      "grad_norm": 0.16290684883919596,
+      "learning_rate": 1.6779887891701068e-06,
+      "loss": 0.426,
+      "step": 2226
+    },
+    {
+      "epoch": 2.1962040916933696,
+      "grad_norm": 0.1562685873603517,
+      "learning_rate": 1.6741249953353434e-06,
+      "loss": 0.3952,
+      "step": 2227
+    },
+    {
+      "epoch": 2.197190041902884,
+      "grad_norm": 0.15898611648713112,
+      "learning_rate": 1.6702647604027178e-06,
+      "loss": 0.4211,
+      "step": 2228
+    },
+    {
+      "epoch": 2.1981759921123984,
+      "grad_norm": 0.17053474456190576,
+      "learning_rate": 1.6664080885029328e-06,
+      "loss": 0.4076,
+      "step": 2229
+    },
+    {
+      "epoch": 2.1991619423219126,
+      "grad_norm": 0.16834333834852952,
+      "learning_rate": 1.6625549837628773e-06,
+      "loss": 0.4305,
+      "step": 2230
+    },
+    {
+      "epoch": 2.2001478925314273,
+      "grad_norm": 0.15729964613924843,
+      "learning_rate": 1.6587054503056232e-06,
+      "loss": 0.4075,
+      "step": 2231
+    },
+    {
+      "epoch": 2.2011338427409415,
+      "grad_norm": 0.1655216429147605,
+      "learning_rate": 1.654859492250422e-06,
+      "loss": 0.4076,
+      "step": 2232
+    },
+    {
+      "epoch": 2.202119792950456,
+      "grad_norm": 0.16018527329403168,
+      "learning_rate": 1.6510171137126974e-06,
+      "loss": 0.4046,
+      "step": 2233
+    },
+    {
+      "epoch": 2.2031057431599703,
+      "grad_norm": 0.1576112435570535,
+      "learning_rate": 1.647178318804043e-06,
+      "loss": 0.393,
+      "step": 2234
+    },
+    {
+      "epoch": 2.204091693369485,
+      "grad_norm": 0.1563385792405461,
+      "learning_rate": 1.6433431116322235e-06,
+      "loss": 0.4135,
+      "step": 2235
+    },
+    {
+      "epoch": 2.205077643578999,
+      "grad_norm": 0.1657235083932023,
+      "learning_rate": 1.6395114963011538e-06,
+      "loss": 0.4072,
+      "step": 2236
+    },
+    {
+      "epoch": 2.206063593788514,
+      "grad_norm": 0.1600732025809443,
+      "learning_rate": 1.6356834769109114e-06,
+      "loss": 0.4054,
+      "step": 2237
+    },
+    {
+      "epoch": 2.207049543998028,
+      "grad_norm": 0.16036074523541055,
+      "learning_rate": 1.6318590575577293e-06,
+      "loss": 0.4186,
+      "step": 2238
+    },
+    {
+      "epoch": 2.2080354942075426,
+      "grad_norm": 0.15797831064291587,
+      "learning_rate": 1.6280382423339818e-06,
+      "loss": 0.4072,
+      "step": 2239
+    },
+    {
+      "epoch": 2.209021444417057,
+      "grad_norm": 0.15994603084363623,
+      "learning_rate": 1.6242210353281922e-06,
+      "loss": 0.426,
+      "step": 2240
+    },
+    {
+      "epoch": 2.2100073946265715,
+      "grad_norm": 0.1681290966340985,
+      "learning_rate": 1.6204074406250136e-06,
+      "loss": 0.4333,
+      "step": 2241
+    },
+    {
+      "epoch": 2.2109933448360857,
+      "grad_norm": 0.15643437806841407,
+      "learning_rate": 1.6165974623052455e-06,
+      "loss": 0.4048,
+      "step": 2242
+    },
+    {
+      "epoch": 2.2119792950456003,
+      "grad_norm": 0.16355208894238868,
+      "learning_rate": 1.6127911044458106e-06,
+      "loss": 0.4145,
+      "step": 2243
+    },
+    {
+      "epoch": 2.2129652452551145,
+      "grad_norm": 0.1641335483343855,
+      "learning_rate": 1.608988371119758e-06,
+      "loss": 0.4042,
+      "step": 2244
+    },
+    {
+      "epoch": 2.213951195464629,
+      "grad_norm": 0.16396929782208178,
+      "learning_rate": 1.6051892663962593e-06,
+      "loss": 0.4193,
+      "step": 2245
+    },
+    {
+      "epoch": 2.2149371456741433,
+      "grad_norm": 0.16150400403270612,
+      "learning_rate": 1.6013937943406038e-06,
+      "loss": 0.4158,
+      "step": 2246
+    },
+    {
+      "epoch": 2.215923095883658,
+      "grad_norm": 0.16093855664741763,
+      "learning_rate": 1.5976019590141929e-06,
+      "loss": 0.3989,
+      "step": 2247
+    },
+    {
+      "epoch": 2.216909046093172,
+      "grad_norm": 0.16155833977339337,
+      "learning_rate": 1.593813764474536e-06,
+      "loss": 0.3997,
+      "step": 2248
+    },
+    {
+      "epoch": 2.217894996302687,
+      "grad_norm": 0.16088300640877046,
+      "learning_rate": 1.5900292147752462e-06,
+      "loss": 0.4163,
+      "step": 2249
+    },
+    {
+      "epoch": 2.218880946512201,
+      "grad_norm": 0.16078202786819562,
+      "learning_rate": 1.5862483139660413e-06,
+      "loss": 0.4148,
+      "step": 2250
+    },
+    {
+      "epoch": 2.2198668967217157,
+      "grad_norm": 0.15925355651700746,
+      "learning_rate": 1.5824710660927268e-06,
+      "loss": 0.4235,
+      "step": 2251
+    },
+    {
+      "epoch": 2.22085284693123,
+      "grad_norm": 0.16207978965295122,
+      "learning_rate": 1.5786974751972033e-06,
+      "loss": 0.4183,
+      "step": 2252
+    },
+    {
+      "epoch": 2.2218387971407445,
+      "grad_norm": 0.16429033852290004,
+      "learning_rate": 1.5749275453174584e-06,
+      "loss": 0.3981,
+      "step": 2253
+    },
+    {
+      "epoch": 2.2228247473502587,
+      "grad_norm": 0.16587177946082177,
+      "learning_rate": 1.5711612804875632e-06,
+      "loss": 0.4239,
+      "step": 2254
+    },
+    {
+      "epoch": 2.2238106975597733,
+      "grad_norm": 0.16213834725171952,
+      "learning_rate": 1.567398684737666e-06,
+      "loss": 0.4227,
+      "step": 2255
+    },
+    {
+      "epoch": 2.2247966477692875,
+      "grad_norm": 0.16324806110391102,
+      "learning_rate": 1.5636397620939842e-06,
+      "loss": 0.3984,
+      "step": 2256
+    },
+    {
+      "epoch": 2.225782597978802,
+      "grad_norm": 0.16422658299514456,
+      "learning_rate": 1.5598845165788134e-06,
+      "loss": 0.4068,
+      "step": 2257
+    },
+    {
+      "epoch": 2.2267685481883164,
+      "grad_norm": 0.16303998595459007,
+      "learning_rate": 1.5561329522105083e-06,
+      "loss": 0.4065,
+      "step": 2258
+    },
+    {
+      "epoch": 2.227754498397831,
+      "grad_norm": 0.16431900778898972,
+      "learning_rate": 1.5523850730034874e-06,
+      "loss": 0.4177,
+      "step": 2259
+    },
+    {
+      "epoch": 2.228740448607345,
+      "grad_norm": 0.16543433401498112,
+      "learning_rate": 1.5486408829682232e-06,
+      "loss": 0.4224,
+      "step": 2260
+    },
+    {
+      "epoch": 2.22972639881686,
+      "grad_norm": 0.16561195089040817,
+      "learning_rate": 1.5449003861112427e-06,
+      "loss": 0.3978,
+      "step": 2261
+    },
+    {
+      "epoch": 2.230712349026374,
+      "grad_norm": 0.1600352065459141,
+      "learning_rate": 1.5411635864351204e-06,
+      "loss": 0.4025,
+      "step": 2262
+    },
+    {
+      "epoch": 2.2316982992358887,
+      "grad_norm": 0.16229768962021862,
+      "learning_rate": 1.5374304879384744e-06,
+      "loss": 0.4125,
+      "step": 2263
+    },
+    {
+      "epoch": 2.232684249445403,
+      "grad_norm": 0.16295485519278402,
+      "learning_rate": 1.5337010946159609e-06,
+      "loss": 0.4117,
+      "step": 2264
+    },
+    {
+      "epoch": 2.2336701996549175,
+      "grad_norm": 0.1663620317984901,
+      "learning_rate": 1.5299754104582765e-06,
+      "loss": 0.4214,
+      "step": 2265
+    },
+    {
+      "epoch": 2.2346561498644317,
+      "grad_norm": 0.16820707279695954,
+      "learning_rate": 1.526253439452144e-06,
+      "loss": 0.4194,
+      "step": 2266
+    },
+    {
+      "epoch": 2.2356421000739464,
+      "grad_norm": 0.16728411638631385,
+      "learning_rate": 1.5225351855803117e-06,
+      "loss": 0.4057,
+      "step": 2267
+    },
+    {
+      "epoch": 2.2366280502834606,
+      "grad_norm": 0.17620979782236681,
+      "learning_rate": 1.5188206528215522e-06,
+      "loss": 0.4107,
+      "step": 2268
+    },
+    {
+      "epoch": 2.237614000492975,
+      "grad_norm": 0.159934923213865,
+      "learning_rate": 1.5151098451506596e-06,
+      "loss": 0.4122,
+      "step": 2269
+    },
+    {
+      "epoch": 2.2385999507024894,
+      "grad_norm": 0.16383742998598416,
+      "learning_rate": 1.5114027665384384e-06,
+      "loss": 0.4096,
+      "step": 2270
+    },
+    {
+      "epoch": 2.239585900912004,
+      "grad_norm": 0.16534238981185112,
+      "learning_rate": 1.5076994209517038e-06,
+      "loss": 0.4185,
+      "step": 2271
+    },
+    {
+      "epoch": 2.2405718511215182,
+      "grad_norm": 0.16226097304859072,
+      "learning_rate": 1.5039998123532717e-06,
+      "loss": 0.392,
+      "step": 2272
+    },
+    {
+      "epoch": 2.241557801331033,
+      "grad_norm": 0.16941902784767632,
+      "learning_rate": 1.500303944701968e-06,
+      "loss": 0.4173,
+      "step": 2273
+    },
+    {
+      "epoch": 2.242543751540547,
+      "grad_norm": 0.16557680158057791,
+      "learning_rate": 1.4966118219526099e-06,
+      "loss": 0.42,
+      "step": 2274
+    },
+    {
+      "epoch": 2.2435297017500617,
+      "grad_norm": 0.16206832978749852,
+      "learning_rate": 1.4929234480560078e-06,
+      "loss": 0.4146,
+      "step": 2275
+    },
+    {
+      "epoch": 2.244515651959576,
+      "grad_norm": 0.16827476886247675,
+      "learning_rate": 1.4892388269589615e-06,
+      "loss": 0.3898,
+      "step": 2276
+    },
+    {
+      "epoch": 2.2455016021690906,
+      "grad_norm": 0.16194306645525028,
+      "learning_rate": 1.4855579626042542e-06,
+      "loss": 0.3905,
+      "step": 2277
+    },
+    {
+      "epoch": 2.2464875523786048,
+      "grad_norm": 0.16473436355046234,
+      "learning_rate": 1.481880858930651e-06,
+      "loss": 0.406,
+      "step": 2278
+    },
+    {
+      "epoch": 2.2474735025881194,
+      "grad_norm": 0.17270595927025267,
+      "learning_rate": 1.47820751987289e-06,
+      "loss": 0.4133,
+      "step": 2279
+    },
+    {
+      "epoch": 2.2484594527976336,
+      "grad_norm": 0.16459987916542002,
+      "learning_rate": 1.4745379493616817e-06,
+      "loss": 0.4143,
+      "step": 2280
+    },
+    {
+      "epoch": 2.2494454030071482,
+      "grad_norm": 0.16364511705619886,
+      "learning_rate": 1.4708721513237096e-06,
+      "loss": 0.4154,
+      "step": 2281
+    },
+    {
+      "epoch": 2.2504313532166624,
+      "grad_norm": 0.1588633085001226,
+      "learning_rate": 1.4672101296816099e-06,
+      "loss": 0.4105,
+      "step": 2282
+    },
+    {
+      "epoch": 2.251417303426177,
+      "grad_norm": 0.15929228670510678,
+      "learning_rate": 1.4635518883539846e-06,
+      "loss": 0.4025,
+      "step": 2283
+    },
+    {
+      "epoch": 2.2524032536356913,
+      "grad_norm": 0.16179783135092632,
+      "learning_rate": 1.4598974312553915e-06,
+      "loss": 0.4163,
+      "step": 2284
+    },
+    {
+      "epoch": 2.253389203845206,
+      "grad_norm": 0.1614449546919737,
+      "learning_rate": 1.4562467622963367e-06,
+      "loss": 0.3943,
+      "step": 2285
+    },
+    {
+      "epoch": 2.25437515405472,
+      "grad_norm": 0.16651728964995471,
+      "learning_rate": 1.4525998853832729e-06,
+      "loss": 0.4216,
+      "step": 2286
+    },
+    {
+      "epoch": 2.2553611042642348,
+      "grad_norm": 0.18272127165647753,
+      "learning_rate": 1.4489568044185914e-06,
+      "loss": 0.413,
+      "step": 2287
+    },
+    {
+      "epoch": 2.256347054473749,
+      "grad_norm": 0.1643655873287094,
+      "learning_rate": 1.4453175233006295e-06,
+      "loss": 0.4149,
+      "step": 2288
+    },
+    {
+      "epoch": 2.2573330046832636,
+      "grad_norm": 0.1716570649131217,
+      "learning_rate": 1.441682045923653e-06,
+      "loss": 0.3974,
+      "step": 2289
+    },
+    {
+      "epoch": 2.258318954892778,
+      "grad_norm": 0.19409339613015558,
+      "learning_rate": 1.4380503761778585e-06,
+      "loss": 0.3761,
+      "step": 2290
+    },
+    {
+      "epoch": 2.2593049051022924,
+      "grad_norm": 0.16359834650509225,
+      "learning_rate": 1.4344225179493687e-06,
+      "loss": 0.4342,
+      "step": 2291
+    },
+    {
+      "epoch": 2.2602908553118066,
+      "grad_norm": 0.1741665404298848,
+      "learning_rate": 1.4307984751202274e-06,
+      "loss": 0.3925,
+      "step": 2292
+    },
+    {
+      "epoch": 2.2612768055213213,
+      "grad_norm": 0.1614225854140192,
+      "learning_rate": 1.4271782515683952e-06,
+      "loss": 0.4044,
+      "step": 2293
+    },
+    {
+      "epoch": 2.2622627557308355,
+      "grad_norm": 0.16395102476888754,
+      "learning_rate": 1.4235618511677462e-06,
+      "loss": 0.4039,
+      "step": 2294
+    },
+    {
+      "epoch": 2.26324870594035,
+      "grad_norm": 0.16325766141781928,
+      "learning_rate": 1.4199492777880624e-06,
+      "loss": 0.4174,
+      "step": 2295
+    },
+    {
+      "epoch": 2.2642346561498643,
+      "grad_norm": 0.1596825751015749,
+      "learning_rate": 1.4163405352950365e-06,
+      "loss": 0.4171,
+      "step": 2296
+    },
+    {
+      "epoch": 2.265220606359379,
+      "grad_norm": 0.15904624178418814,
+      "learning_rate": 1.412735627550253e-06,
+      "loss": 0.3946,
+      "step": 2297
+    },
+    {
+      "epoch": 2.266206556568893,
+      "grad_norm": 0.16666928908687545,
+      "learning_rate": 1.4091345584111976e-06,
+      "loss": 0.4212,
+      "step": 2298
+    },
+    {
+      "epoch": 2.267192506778408,
+      "grad_norm": 0.1581487250365793,
+      "learning_rate": 1.405537331731247e-06,
+      "loss": 0.4085,
+      "step": 2299
+    },
+    {
+      "epoch": 2.268178456987922,
+      "grad_norm": 0.18782944877358512,
+      "learning_rate": 1.4019439513596705e-06,
+      "loss": 0.4176,
+      "step": 2300
+    },
+    {
+      "epoch": 2.2691644071974366,
+      "grad_norm": 0.16578694247998996,
+      "learning_rate": 1.3983544211416184e-06,
+      "loss": 0.4135,
+      "step": 2301
+    },
+    {
+      "epoch": 2.270150357406951,
+      "grad_norm": 0.16377742221519453,
+      "learning_rate": 1.3947687449181158e-06,
+      "loss": 0.4303,
+      "step": 2302
+    },
+    {
+      "epoch": 2.2711363076164655,
+      "grad_norm": 0.16748897978969077,
+      "learning_rate": 1.391186926526074e-06,
+      "loss": 0.4281,
+      "step": 2303
+    },
+    {
+      "epoch": 2.2721222578259797,
+      "grad_norm": 0.16299536900408998,
+      "learning_rate": 1.3876089697982704e-06,
+      "loss": 0.4079,
+      "step": 2304
+    },
+    {
+      "epoch": 2.2731082080354943,
+      "grad_norm": 0.18379707681905336,
+      "learning_rate": 1.3840348785633494e-06,
+      "loss": 0.4274,
+      "step": 2305
+    },
+    {
+      "epoch": 2.2740941582450085,
+      "grad_norm": 0.1670828207958332,
+      "learning_rate": 1.3804646566458225e-06,
+      "loss": 0.4219,
+      "step": 2306
+    },
+    {
+      "epoch": 2.275080108454523,
+      "grad_norm": 0.16235536756139166,
+      "learning_rate": 1.3768983078660569e-06,
+      "loss": 0.4287,
+      "step": 2307
+    },
+    {
+      "epoch": 2.2760660586640373,
+      "grad_norm": 0.16478870974164742,
+      "learning_rate": 1.3733358360402788e-06,
+      "loss": 0.4162,
+      "step": 2308
+    },
+    {
+      "epoch": 2.277052008873552,
+      "grad_norm": 0.15814873318539266,
+      "learning_rate": 1.3697772449805635e-06,
+      "loss": 0.3941,
+      "step": 2309
+    },
+    {
+      "epoch": 2.278037959083066,
+      "grad_norm": 0.15759543961241237,
+      "learning_rate": 1.366222538494833e-06,
+      "loss": 0.4147,
+      "step": 2310
+    },
+    {
+      "epoch": 2.279023909292581,
+      "grad_norm": 0.15750970138351825,
+      "learning_rate": 1.362671720386859e-06,
+      "loss": 0.4082,
+      "step": 2311
+    },
+    {
+      "epoch": 2.280009859502095,
+      "grad_norm": 0.16639281028239625,
+      "learning_rate": 1.3591247944562424e-06,
+      "loss": 0.4033,
+      "step": 2312
+    },
+    {
+      "epoch": 2.2809958097116096,
+      "grad_norm": 0.16318389717975754,
+      "learning_rate": 1.3555817644984259e-06,
+      "loss": 0.4121,
+      "step": 2313
+    },
+    {
+      "epoch": 2.281981759921124,
+      "grad_norm": 0.1600569608278272,
+      "learning_rate": 1.3520426343046794e-06,
+      "loss": 0.4154,
+      "step": 2314
+    },
+    {
+      "epoch": 2.2829677101306385,
+      "grad_norm": 0.1579638199323022,
+      "learning_rate": 1.3485074076621063e-06,
+      "loss": 0.4025,
+      "step": 2315
+    },
+    {
+      "epoch": 2.2839536603401527,
+      "grad_norm": 0.15864404733099394,
+      "learning_rate": 1.3449760883536266e-06,
+      "loss": 0.4196,
+      "step": 2316
+    },
+    {
+      "epoch": 2.2849396105496673,
+      "grad_norm": 0.16399490747054,
+      "learning_rate": 1.341448680157979e-06,
+      "loss": 0.4142,
+      "step": 2317
+    },
+    {
+      "epoch": 2.2859255607591815,
+      "grad_norm": 0.16011740107155603,
+      "learning_rate": 1.3379251868497217e-06,
+      "loss": 0.4079,
+      "step": 2318
+    },
+    {
+      "epoch": 2.286911510968696,
+      "grad_norm": 0.16148519434557543,
+      "learning_rate": 1.334405612199221e-06,
+      "loss": 0.4111,
+      "step": 2319
+    },
+    {
+      "epoch": 2.2878974611782104,
+      "grad_norm": 0.15759777608133907,
+      "learning_rate": 1.3308899599726493e-06,
+      "loss": 0.4151,
+      "step": 2320
+    },
+    {
+      "epoch": 2.288883411387725,
+      "grad_norm": 0.16955653152412,
+      "learning_rate": 1.3273782339319835e-06,
+      "loss": 0.4172,
+      "step": 2321
+    },
+    {
+      "epoch": 2.289869361597239,
+      "grad_norm": 0.1644435979901736,
+      "learning_rate": 1.3238704378349987e-06,
+      "loss": 0.4416,
+      "step": 2322
+    },
+    {
+      "epoch": 2.290855311806754,
+      "grad_norm": 0.16345034177638715,
+      "learning_rate": 1.320366575435263e-06,
+      "loss": 0.4164,
+      "step": 2323
+    },
+    {
+      "epoch": 2.291841262016268,
+      "grad_norm": 0.16415191454126113,
+      "learning_rate": 1.3168666504821375e-06,
+      "loss": 0.4098,
+      "step": 2324
+    },
+    {
+      "epoch": 2.2928272122257827,
+      "grad_norm": 0.16565438768318477,
+      "learning_rate": 1.3133706667207697e-06,
+      "loss": 0.422,
+      "step": 2325
+    },
+    {
+      "epoch": 2.293813162435297,
+      "grad_norm": 0.15995448712585397,
+      "learning_rate": 1.3098786278920877e-06,
+      "loss": 0.4154,
+      "step": 2326
+    },
+    {
+      "epoch": 2.2947991126448115,
+      "grad_norm": 0.16198130499563007,
+      "learning_rate": 1.3063905377328006e-06,
+      "loss": 0.4279,
+      "step": 2327
+    },
+    {
+      "epoch": 2.2957850628543257,
+      "grad_norm": 0.1617429066817796,
+      "learning_rate": 1.3029063999753916e-06,
+      "loss": 0.4205,
+      "step": 2328
+    },
+    {
+      "epoch": 2.2967710130638403,
+      "grad_norm": 0.1635844701720503,
+      "learning_rate": 1.2994262183481121e-06,
+      "loss": 0.4173,
+      "step": 2329
+    },
+    {
+      "epoch": 2.2977569632733545,
+      "grad_norm": 0.15800385788487895,
+      "learning_rate": 1.2959499965749855e-06,
+      "loss": 0.3905,
+      "step": 2330
+    },
+    {
+      "epoch": 2.298742913482869,
+      "grad_norm": 0.16250204875623522,
+      "learning_rate": 1.2924777383757947e-06,
+      "loss": 0.4118,
+      "step": 2331
+    },
+    {
+      "epoch": 2.2997288636923834,
+      "grad_norm": 0.16294872877278363,
+      "learning_rate": 1.2890094474660785e-06,
+      "loss": 0.4215,
+      "step": 2332
+    },
+    {
+      "epoch": 2.300714813901898,
+      "grad_norm": 0.15911908618348297,
+      "learning_rate": 1.2855451275571335e-06,
+      "loss": 0.3932,
+      "step": 2333
+    },
+    {
+      "epoch": 2.3017007641114122,
+      "grad_norm": 0.16051854421822973,
+      "learning_rate": 1.2820847823560095e-06,
+      "loss": 0.4131,
+      "step": 2334
+    },
+    {
+      "epoch": 2.302686714320927,
+      "grad_norm": 0.16169563962111952,
+      "learning_rate": 1.2786284155654993e-06,
+      "loss": 0.433,
+      "step": 2335
+    },
+    {
+      "epoch": 2.303672664530441,
+      "grad_norm": 0.1610602524410069,
+      "learning_rate": 1.2751760308841393e-06,
+      "loss": 0.4312,
+      "step": 2336
+    },
+    {
+      "epoch": 2.3046586147399557,
+      "grad_norm": 0.16258865600814715,
+      "learning_rate": 1.2717276320062055e-06,
+      "loss": 0.4178,
+      "step": 2337
+    },
+    {
+      "epoch": 2.30564456494947,
+      "grad_norm": 0.16625316687523686,
+      "learning_rate": 1.2682832226217085e-06,
+      "loss": 0.4276,
+      "step": 2338
+    },
+    {
+      "epoch": 2.3066305151589845,
+      "grad_norm": 0.17782087539321642,
+      "learning_rate": 1.2648428064163898e-06,
+      "loss": 0.4058,
+      "step": 2339
+    },
+    {
+      "epoch": 2.3076164653684987,
+      "grad_norm": 0.16349501225562255,
+      "learning_rate": 1.261406387071718e-06,
+      "loss": 0.4248,
+      "step": 2340
+    },
+    {
+      "epoch": 2.3086024155780134,
+      "grad_norm": 0.18761252560626943,
+      "learning_rate": 1.257973968264885e-06,
+      "loss": 0.3802,
+      "step": 2341
+    },
+    {
+      "epoch": 2.3095883657875276,
+      "grad_norm": 0.16517967277010326,
+      "learning_rate": 1.2545455536688022e-06,
+      "loss": 0.4261,
+      "step": 2342
+    },
+    {
+      "epoch": 2.310574315997042,
+      "grad_norm": 0.15863580873816105,
+      "learning_rate": 1.2511211469520945e-06,
+      "loss": 0.3891,
+      "step": 2343
+    },
+    {
+      "epoch": 2.3115602662065564,
+      "grad_norm": 0.16029761268638454,
+      "learning_rate": 1.2477007517791007e-06,
+      "loss": 0.4195,
+      "step": 2344
+    },
+    {
+      "epoch": 2.312546216416071,
+      "grad_norm": 0.16565382525985026,
+      "learning_rate": 1.2442843718098635e-06,
+      "loss": 0.411,
+      "step": 2345
+    },
+    {
+      "epoch": 2.3135321666255853,
+      "grad_norm": 0.15986114270844246,
+      "learning_rate": 1.2408720107001343e-06,
+      "loss": 0.4082,
+      "step": 2346
+    },
+    {
+      "epoch": 2.3145181168351,
+      "grad_norm": 0.16280923514839396,
+      "learning_rate": 1.237463672101361e-06,
+      "loss": 0.4141,
+      "step": 2347
+    },
+    {
+      "epoch": 2.315504067044614,
+      "grad_norm": 0.1604111045926094,
+      "learning_rate": 1.2340593596606832e-06,
+      "loss": 0.3961,
+      "step": 2348
+    },
+    {
+      "epoch": 2.3164900172541287,
+      "grad_norm": 0.16205921559293093,
+      "learning_rate": 1.2306590770209393e-06,
+      "loss": 0.4351,
+      "step": 2349
+    },
+    {
+      "epoch": 2.317475967463643,
+      "grad_norm": 0.17346077650464367,
+      "learning_rate": 1.2272628278206521e-06,
+      "loss": 0.4266,
+      "step": 2350
+    },
+    {
+      "epoch": 2.3184619176731576,
+      "grad_norm": 0.1689247568840012,
+      "learning_rate": 1.2238706156940284e-06,
+      "loss": 0.413,
+      "step": 2351
+    },
+    {
+      "epoch": 2.3194478678826718,
+      "grad_norm": 0.35044658063633505,
+      "learning_rate": 1.220482444270955e-06,
+      "loss": 0.3969,
+      "step": 2352
+    },
+    {
+      "epoch": 2.3204338180921864,
+      "grad_norm": 0.1681692999703932,
+      "learning_rate": 1.2170983171769963e-06,
+      "loss": 0.4279,
+      "step": 2353
+    },
+    {
+      "epoch": 2.3214197683017006,
+      "grad_norm": 0.1639903546130607,
+      "learning_rate": 1.2137182380333867e-06,
+      "loss": 0.4029,
+      "step": 2354
+    },
+    {
+      "epoch": 2.3224057185112152,
+      "grad_norm": 0.1657581841212683,
+      "learning_rate": 1.2103422104570311e-06,
+      "loss": 0.4144,
+      "step": 2355
+    },
+    {
+      "epoch": 2.3233916687207294,
+      "grad_norm": 0.15761773790631234,
+      "learning_rate": 1.2069702380604958e-06,
+      "loss": 0.4151,
+      "step": 2356
+    },
+    {
+      "epoch": 2.324377618930244,
+      "grad_norm": 0.16581608724165262,
+      "learning_rate": 1.2036023244520157e-06,
+      "loss": 0.4132,
+      "step": 2357
+    },
+    {
+      "epoch": 2.3253635691397583,
+      "grad_norm": 0.16660329033788626,
+      "learning_rate": 1.200238473235472e-06,
+      "loss": 0.3983,
+      "step": 2358
+    },
+    {
+      "epoch": 2.326349519349273,
+      "grad_norm": 0.16813796837382863,
+      "learning_rate": 1.1968786880104049e-06,
+      "loss": 0.417,
+      "step": 2359
+    },
+    {
+      "epoch": 2.327335469558787,
+      "grad_norm": 0.1606712490218808,
+      "learning_rate": 1.193522972372002e-06,
+      "loss": 0.4141,
+      "step": 2360
+    },
+    {
+      "epoch": 2.3283214197683018,
+      "grad_norm": 0.1662513472911015,
+      "learning_rate": 1.1901713299110995e-06,
+      "loss": 0.4227,
+      "step": 2361
+    },
+    {
+      "epoch": 2.329307369977816,
+      "grad_norm": 0.1701130339328463,
+      "learning_rate": 1.1868237642141723e-06,
+      "loss": 0.4232,
+      "step": 2362
+    },
+    {
+      "epoch": 2.3302933201873306,
+      "grad_norm": 0.17146127994694974,
+      "learning_rate": 1.1834802788633288e-06,
+      "loss": 0.4101,
+      "step": 2363
+    },
+    {
+      "epoch": 2.331279270396845,
+      "grad_norm": 0.15743290116696995,
+      "learning_rate": 1.1801408774363188e-06,
+      "loss": 0.4103,
+      "step": 2364
+    },
+    {
+      "epoch": 2.3322652206063594,
+      "grad_norm": 0.1568591856904107,
+      "learning_rate": 1.1768055635065184e-06,
+      "loss": 0.4191,
+      "step": 2365
+    },
+    {
+      "epoch": 2.3332511708158736,
+      "grad_norm": 0.16019347204307025,
+      "learning_rate": 1.1734743406429295e-06,
+      "loss": 0.415,
+      "step": 2366
+    },
+    {
+      "epoch": 2.3342371210253883,
+      "grad_norm": 0.16613274290188823,
+      "learning_rate": 1.1701472124101765e-06,
+      "loss": 0.4224,
+      "step": 2367
+    },
+    {
+      "epoch": 2.3352230712349025,
+      "grad_norm": 0.1725775485591128,
+      "learning_rate": 1.1668241823685028e-06,
+      "loss": 0.4063,
+      "step": 2368
+    },
+    {
+      "epoch": 2.336209021444417,
+      "grad_norm": 0.19590817684839829,
+      "learning_rate": 1.1635052540737668e-06,
+      "loss": 0.4092,
+      "step": 2369
+    },
+    {
+      "epoch": 2.3371949716539313,
+      "grad_norm": 0.160213657045184,
+      "learning_rate": 1.1601904310774364e-06,
+      "loss": 0.4183,
+      "step": 2370
+    },
+    {
+      "epoch": 2.338180921863446,
+      "grad_norm": 0.1650783394041844,
+      "learning_rate": 1.1568797169265878e-06,
+      "loss": 0.4101,
+      "step": 2371
+    },
+    {
+      "epoch": 2.33916687207296,
+      "grad_norm": 0.16119554712600798,
+      "learning_rate": 1.1535731151638997e-06,
+      "loss": 0.4064,
+      "step": 2372
+    },
+    {
+      "epoch": 2.340152822282475,
+      "grad_norm": 0.1879428160296416,
+      "learning_rate": 1.1502706293276504e-06,
+      "loss": 0.4212,
+      "step": 2373
+    },
+    {
+      "epoch": 2.341138772491989,
+      "grad_norm": 0.1629396282658936,
+      "learning_rate": 1.1469722629517156e-06,
+      "loss": 0.4216,
+      "step": 2374
+    },
+    {
+      "epoch": 2.3421247227015036,
+      "grad_norm": 0.1619454303955732,
+      "learning_rate": 1.1436780195655583e-06,
+      "loss": 0.3997,
+      "step": 2375
+    },
+    {
+      "epoch": 2.343110672911018,
+      "grad_norm": 0.15899330964730712,
+      "learning_rate": 1.1403879026942361e-06,
+      "loss": 0.4201,
+      "step": 2376
+    },
+    {
+      "epoch": 2.3440966231205325,
+      "grad_norm": 0.16546029939437998,
+      "learning_rate": 1.137101915858388e-06,
+      "loss": 0.4069,
+      "step": 2377
+    },
+    {
+      "epoch": 2.3450825733300467,
+      "grad_norm": 0.15966260774396165,
+      "learning_rate": 1.13382006257423e-06,
+      "loss": 0.4152,
+      "step": 2378
+    },
+    {
+      "epoch": 2.3460685235395613,
+      "grad_norm": 0.1628680329736944,
+      "learning_rate": 1.130542346353558e-06,
+      "loss": 0.4191,
+      "step": 2379
+    },
+    {
+      "epoch": 2.3470544737490755,
+      "grad_norm": 0.18130856121973396,
+      "learning_rate": 1.127268770703745e-06,
+      "loss": 0.4224,
+      "step": 2380
+    },
+    {
+      "epoch": 2.34804042395859,
+      "grad_norm": 0.15908995889307495,
+      "learning_rate": 1.1239993391277264e-06,
+      "loss": 0.4286,
+      "step": 2381
+    },
+    {
+      "epoch": 2.3490263741681043,
+      "grad_norm": 0.1595524071803773,
+      "learning_rate": 1.1207340551240076e-06,
+      "loss": 0.411,
+      "step": 2382
+    },
+    {
+      "epoch": 2.350012324377619,
+      "grad_norm": 0.16459379825706655,
+      "learning_rate": 1.1174729221866532e-06,
+      "loss": 0.4192,
+      "step": 2383
+    },
+    {
+      "epoch": 2.350998274587133,
+      "grad_norm": 0.16973630430457837,
+      "learning_rate": 1.1142159438052886e-06,
+      "loss": 0.4406,
+      "step": 2384
+    },
+    {
+      "epoch": 2.351984224796648,
+      "grad_norm": 0.16294530645656632,
+      "learning_rate": 1.1109631234650903e-06,
+      "loss": 0.3993,
+      "step": 2385
+    },
+    {
+      "epoch": 2.352970175006162,
+      "grad_norm": 0.16294863825764402,
+      "learning_rate": 1.107714464646789e-06,
+      "loss": 0.4092,
+      "step": 2386
+    },
+    {
+      "epoch": 2.3539561252156767,
+      "grad_norm": 0.16891054832785016,
+      "learning_rate": 1.1044699708266594e-06,
+      "loss": 0.4206,
+      "step": 2387
+    },
+    {
+      "epoch": 2.354942075425191,
+      "grad_norm": 0.1692897374545682,
+      "learning_rate": 1.1012296454765208e-06,
+      "loss": 0.4202,
+      "step": 2388
+    },
+    {
+      "epoch": 2.3559280256347055,
+      "grad_norm": 0.16625081299129083,
+      "learning_rate": 1.0979934920637314e-06,
+      "loss": 0.4097,
+      "step": 2389
+    },
+    {
+      "epoch": 2.3569139758442197,
+      "grad_norm": 0.16511966654296223,
+      "learning_rate": 1.0947615140511858e-06,
+      "loss": 0.4032,
+      "step": 2390
+    },
+    {
+      "epoch": 2.3578999260537343,
+      "grad_norm": 0.16264373202500912,
+      "learning_rate": 1.0915337148973088e-06,
+      "loss": 0.4199,
+      "step": 2391
+    },
+    {
+      "epoch": 2.3588858762632485,
+      "grad_norm": 0.1689305005673266,
+      "learning_rate": 1.088310098056059e-06,
+      "loss": 0.4006,
+      "step": 2392
+    },
+    {
+      "epoch": 2.359871826472763,
+      "grad_norm": 0.16546899130059672,
+      "learning_rate": 1.0850906669769113e-06,
+      "loss": 0.4244,
+      "step": 2393
+    },
+    {
+      "epoch": 2.3608577766822774,
+      "grad_norm": 0.15428546428468326,
+      "learning_rate": 1.0818754251048664e-06,
+      "loss": 0.3785,
+      "step": 2394
+    },
+    {
+      "epoch": 2.361843726891792,
+      "grad_norm": 0.15805192109688979,
+      "learning_rate": 1.0786643758804444e-06,
+      "loss": 0.4169,
+      "step": 2395
+    },
+    {
+      "epoch": 2.362829677101306,
+      "grad_norm": 0.1560837158001152,
+      "learning_rate": 1.075457522739675e-06,
+      "loss": 0.4083,
+      "step": 2396
+    },
+    {
+      "epoch": 2.363815627310821,
+      "grad_norm": 0.1626728416196687,
+      "learning_rate": 1.072254869114101e-06,
+      "loss": 0.4255,
+      "step": 2397
+    },
+    {
+      "epoch": 2.364801577520335,
+      "grad_norm": 0.16345611835914114,
+      "learning_rate": 1.0690564184307645e-06,
+      "loss": 0.423,
+      "step": 2398
+    },
+    {
+      "epoch": 2.3657875277298497,
+      "grad_norm": 0.16007046777382433,
+      "learning_rate": 1.0658621741122205e-06,
+      "loss": 0.4189,
+      "step": 2399
+    },
+    {
+      "epoch": 2.366773477939364,
+      "grad_norm": 0.15778022472510203,
+      "learning_rate": 1.062672139576516e-06,
+      "loss": 0.4147,
+      "step": 2400
+    },
+    {
+      "epoch": 2.3677594281488785,
+      "grad_norm": 0.16476560246321204,
+      "learning_rate": 1.059486318237195e-06,
+      "loss": 0.4369,
+      "step": 2401
+    },
+    {
+      "epoch": 2.3687453783583927,
+      "grad_norm": 0.1614508122516947,
+      "learning_rate": 1.0563047135032928e-06,
+      "loss": 0.419,
+      "step": 2402
+    },
+    {
+      "epoch": 2.3697313285679074,
+      "grad_norm": 0.15979009015839116,
+      "learning_rate": 1.0531273287793336e-06,
+      "loss": 0.4016,
+      "step": 2403
+    },
+    {
+      "epoch": 2.3707172787774216,
+      "grad_norm": 0.19544947099284235,
+      "learning_rate": 1.0499541674653251e-06,
+      "loss": 0.4162,
+      "step": 2404
+    },
+    {
+      "epoch": 2.371703228986936,
+      "grad_norm": 0.1671507504998902,
+      "learning_rate": 1.0467852329567558e-06,
+      "loss": 0.4219,
+      "step": 2405
+    },
+    {
+      "epoch": 2.3726891791964504,
+      "grad_norm": 0.15983792455754875,
+      "learning_rate": 1.0436205286445893e-06,
+      "loss": 0.4164,
+      "step": 2406
+    },
+    {
+      "epoch": 2.373675129405965,
+      "grad_norm": 0.1565821981045399,
+      "learning_rate": 1.0404600579152702e-06,
+      "loss": 0.3978,
+      "step": 2407
+    },
+    {
+      "epoch": 2.3746610796154792,
+      "grad_norm": 0.16628224983624315,
+      "learning_rate": 1.0373038241507017e-06,
+      "loss": 0.4066,
+      "step": 2408
+    },
+    {
+      "epoch": 2.375647029824994,
+      "grad_norm": 0.1548084367651464,
+      "learning_rate": 1.03415183072826e-06,
+      "loss": 0.4055,
+      "step": 2409
+    },
+    {
+      "epoch": 2.376632980034508,
+      "grad_norm": 0.17412370535395244,
+      "learning_rate": 1.031004081020785e-06,
+      "loss": 0.4321,
+      "step": 2410
+    },
+    {
+      "epoch": 2.3776189302440227,
+      "grad_norm": 0.16819021214759913,
+      "learning_rate": 1.0278605783965712e-06,
+      "loss": 0.4263,
+      "step": 2411
+    },
+    {
+      "epoch": 2.378604880453537,
+      "grad_norm": 0.16241930999439794,
+      "learning_rate": 1.0247213262193728e-06,
+      "loss": 0.4172,
+      "step": 2412
+    },
+    {
+      "epoch": 2.3795908306630515,
+      "grad_norm": 0.18316525935336245,
+      "learning_rate": 1.021586327848389e-06,
+      "loss": 0.4102,
+      "step": 2413
+    },
+    {
+      "epoch": 2.3805767808725657,
+      "grad_norm": 0.16283384740375167,
+      "learning_rate": 1.018455586638275e-06,
+      "loss": 0.4209,
+      "step": 2414
+    },
+    {
+      "epoch": 2.3815627310820804,
+      "grad_norm": 0.15909239037933473,
+      "learning_rate": 1.0153291059391269e-06,
+      "loss": 0.4025,
+      "step": 2415
+    },
+    {
+      "epoch": 2.3825486812915946,
+      "grad_norm": 0.16703744818586627,
+      "learning_rate": 1.012206889096481e-06,
+      "loss": 0.4099,
+      "step": 2416
+    },
+    {
+      "epoch": 2.3835346315011092,
+      "grad_norm": 0.16602750313595738,
+      "learning_rate": 1.009088939451312e-06,
+      "loss": 0.4203,
+      "step": 2417
+    },
+    {
+      "epoch": 2.3845205817106234,
+      "grad_norm": 0.1602146542369932,
+      "learning_rate": 1.0059752603400291e-06,
+      "loss": 0.3774,
+      "step": 2418
+    },
+    {
+      "epoch": 2.385506531920138,
+      "grad_norm": 0.15588894044415524,
+      "learning_rate": 1.0028658550944703e-06,
+      "loss": 0.3932,
+      "step": 2419
+    },
+    {
+      "epoch": 2.3864924821296523,
+      "grad_norm": 0.16134547972178684,
+      "learning_rate": 9.997607270419018e-07,
+      "loss": 0.4115,
+      "step": 2420
+    },
+    {
+      "epoch": 2.387478432339167,
+      "grad_norm": 0.18366412150660394,
+      "learning_rate": 9.96659879505011e-07,
+      "loss": 0.4022,
+      "step": 2421
+    },
+    {
+      "epoch": 2.388464382548681,
+      "grad_norm": 0.15947815873855078,
+      "learning_rate": 9.935633158019087e-07,
+      "loss": 0.3991,
+      "step": 2422
+    },
+    {
+      "epoch": 2.3894503327581957,
+      "grad_norm": 0.16201163898151466,
+      "learning_rate": 9.90471039246116e-07,
+      "loss": 0.4258,
+      "step": 2423
+    },
+    {
+      "epoch": 2.39043628296771,
+      "grad_norm": 0.642697359816267,
+      "learning_rate": 9.873830531465711e-07,
+      "loss": 0.4156,
+      "step": 2424
+    },
+    {
+      "epoch": 2.3914222331772246,
+      "grad_norm": 0.15781129262500931,
+      "learning_rate": 9.842993608076174e-07,
+      "loss": 0.4077,
+      "step": 2425
+    },
+    {
+      "epoch": 2.392408183386739,
+      "grad_norm": 0.16218024963353128,
+      "learning_rate": 9.812199655290095e-07,
+      "loss": 0.4258,
+      "step": 2426
+    },
+    {
+      "epoch": 2.3933941335962534,
+      "grad_norm": 0.16559005058587487,
+      "learning_rate": 9.781448706058983e-07,
+      "loss": 0.412,
+      "step": 2427
+    },
+    {
+      "epoch": 2.3943800838057676,
+      "grad_norm": 0.16040452075409298,
+      "learning_rate": 9.75074079328835e-07,
+      "loss": 0.4281,
+      "step": 2428
+    },
+    {
+      "epoch": 2.3953660340152823,
+      "grad_norm": 0.15969965446120735,
+      "learning_rate": 9.720075949837659e-07,
+      "loss": 0.4102,
+      "step": 2429
+    },
+    {
+      "epoch": 2.396351984224797,
+      "grad_norm": 0.1760116297111136,
+      "learning_rate": 9.689454208520276e-07,
+      "loss": 0.4295,
+      "step": 2430
+    },
+    {
+      "epoch": 2.397337934434311,
+      "grad_norm": 0.16506477568007438,
+      "learning_rate": 9.658875602103461e-07,
+      "loss": 0.4128,
+      "step": 2431
+    },
+    {
+      "epoch": 2.3983238846438253,
+      "grad_norm": 0.1650611715314464,
+      "learning_rate": 9.628340163308304e-07,
+      "loss": 0.4168,
+      "step": 2432
+    },
+    {
+      "epoch": 2.39930983485334,
+      "grad_norm": 0.158898169747212,
+      "learning_rate": 9.5978479248097e-07,
+      "loss": 0.4272,
+      "step": 2433
+    },
+    {
+      "epoch": 2.4002957850628546,
+      "grad_norm": 0.16249408933862527,
+      "learning_rate": 9.567398919236332e-07,
+      "loss": 0.4217,
+      "step": 2434
+    },
+    {
+      "epoch": 2.4012817352723688,
+      "grad_norm": 0.15661362275022744,
+      "learning_rate": 9.536993179170612e-07,
+      "loss": 0.4138,
+      "step": 2435
+    },
+    {
+      "epoch": 2.402267685481883,
+      "grad_norm": 0.16349044834567097,
+      "learning_rate": 9.506630737148642e-07,
+      "loss": 0.4262,
+      "step": 2436
+    },
+    {
+      "epoch": 2.4032536356913976,
+      "grad_norm": 0.15986245087226977,
+      "learning_rate": 9.476311625660228e-07,
+      "loss": 0.4027,
+      "step": 2437
+    },
+    {
+      "epoch": 2.4042395859009122,
+      "grad_norm": 0.16044952949303962,
+      "learning_rate": 9.446035877148785e-07,
+      "loss": 0.4127,
+      "step": 2438
+    },
+    {
+      "epoch": 2.4052255361104264,
+      "grad_norm": 0.16997371196521344,
+      "learning_rate": 9.415803524011313e-07,
+      "loss": 0.4071,
+      "step": 2439
+    },
+    {
+      "epoch": 2.4062114863199406,
+      "grad_norm": 0.16335962168658272,
+      "learning_rate": 9.385614598598386e-07,
+      "loss": 0.4156,
+      "step": 2440
+    },
+    {
+      "epoch": 2.4071974365294553,
+      "grad_norm": 0.161844721602425,
+      "learning_rate": 9.35546913321414e-07,
+      "loss": 0.4213,
+      "step": 2441
+    },
+    {
+      "epoch": 2.40818338673897,
+      "grad_norm": 0.15902052235840378,
+      "learning_rate": 9.325367160116167e-07,
+      "loss": 0.399,
+      "step": 2442
+    },
+    {
+      "epoch": 2.409169336948484,
+      "grad_norm": 0.16331959521354725,
+      "learning_rate": 9.295308711515543e-07,
+      "loss": 0.4076,
+      "step": 2443
+    },
+    {
+      "epoch": 2.4101552871579983,
+      "grad_norm": 0.16060558877947725,
+      "learning_rate": 9.265293819576726e-07,
+      "loss": 0.4244,
+      "step": 2444
+    },
+    {
+      "epoch": 2.411141237367513,
+      "grad_norm": 0.17080526417558123,
+      "learning_rate": 9.235322516417633e-07,
+      "loss": 0.4287,
+      "step": 2445
+    },
+    {
+      "epoch": 2.4121271875770276,
+      "grad_norm": 0.16474200219189994,
+      "learning_rate": 9.205394834109494e-07,
+      "loss": 0.414,
+      "step": 2446
+    },
+    {
+      "epoch": 2.413113137786542,
+      "grad_norm": 0.15616101277976135,
+      "learning_rate": 9.175510804676868e-07,
+      "loss": 0.4115,
+      "step": 2447
+    },
+    {
+      "epoch": 2.414099087996056,
+      "grad_norm": 0.16259371580323215,
+      "learning_rate": 9.145670460097606e-07,
+      "loss": 0.41,
+      "step": 2448
+    },
+    {
+      "epoch": 2.4150850382055706,
+      "grad_norm": 0.1605263138263877,
+      "learning_rate": 9.115873832302818e-07,
+      "loss": 0.403,
+      "step": 2449
+    },
+    {
+      "epoch": 2.4160709884150853,
+      "grad_norm": 0.1640492641760322,
+      "learning_rate": 9.08612095317683e-07,
+      "loss": 0.3997,
+      "step": 2450
+    },
+    {
+      "epoch": 2.4170569386245995,
+      "grad_norm": 0.15425470547068335,
+      "learning_rate": 9.056411854557146e-07,
+      "loss": 0.4098,
+      "step": 2451
+    },
+    {
+      "epoch": 2.4180428888341137,
+      "grad_norm": 0.2353652395628097,
+      "learning_rate": 9.026746568234424e-07,
+      "loss": 0.4212,
+      "step": 2452
+    },
+    {
+      "epoch": 2.4190288390436283,
+      "grad_norm": 0.17285285897048183,
+      "learning_rate": 8.997125125952483e-07,
+      "loss": 0.4289,
+      "step": 2453
+    },
+    {
+      "epoch": 2.420014789253143,
+      "grad_norm": 0.1632310872062539,
+      "learning_rate": 8.967547559408152e-07,
+      "loss": 0.4125,
+      "step": 2454
+    },
+    {
+      "epoch": 2.421000739462657,
+      "grad_norm": 0.15894065204491595,
+      "learning_rate": 8.938013900251346e-07,
+      "loss": 0.4224,
+      "step": 2455
+    },
+    {
+      "epoch": 2.4219866896721713,
+      "grad_norm": 0.16483748236212414,
+      "learning_rate": 8.908524180085021e-07,
+      "loss": 0.4104,
+      "step": 2456
+    },
+    {
+      "epoch": 2.422972639881686,
+      "grad_norm": 0.15904030114263468,
+      "learning_rate": 8.879078430465082e-07,
+      "loss": 0.41,
+      "step": 2457
+    },
+    {
+      "epoch": 2.4239585900912006,
+      "grad_norm": 0.4352207499954632,
+      "learning_rate": 8.849676682900399e-07,
+      "loss": 0.4061,
+      "step": 2458
+    },
+    {
+      "epoch": 2.424944540300715,
+      "grad_norm": 0.16021618921464745,
+      "learning_rate": 8.82031896885272e-07,
+      "loss": 0.4339,
+      "step": 2459
+    },
+    {
+      "epoch": 2.425930490510229,
+      "grad_norm": 0.16891102494389298,
+      "learning_rate": 8.79100531973674e-07,
+      "loss": 0.4089,
+      "step": 2460
+    },
+    {
+      "epoch": 2.4269164407197437,
+      "grad_norm": 0.15956671598032787,
+      "learning_rate": 8.761735766919955e-07,
+      "loss": 0.4087,
+      "step": 2461
+    },
+    {
+      "epoch": 2.4279023909292583,
+      "grad_norm": 0.15840194748311953,
+      "learning_rate": 8.732510341722678e-07,
+      "loss": 0.4186,
+      "step": 2462
+    },
+    {
+      "epoch": 2.4288883411387725,
+      "grad_norm": 0.16173441730914487,
+      "learning_rate": 8.703329075418021e-07,
+      "loss": 0.4245,
+      "step": 2463
+    },
+    {
+      "epoch": 2.4298742913482867,
+      "grad_norm": 0.16185462013620347,
+      "learning_rate": 8.674191999231835e-07,
+      "loss": 0.4431,
+      "step": 2464
+    },
+    {
+      "epoch": 2.4308602415578013,
+      "grad_norm": 0.15926357057296528,
+      "learning_rate": 8.645099144342672e-07,
+      "loss": 0.3945,
+      "step": 2465
+    },
+    {
+      "epoch": 2.431846191767316,
+      "grad_norm": 0.1637089151876667,
+      "learning_rate": 8.616050541881782e-07,
+      "loss": 0.4168,
+      "step": 2466
+    },
+    {
+      "epoch": 2.43283214197683,
+      "grad_norm": 0.1597325067081291,
+      "learning_rate": 8.587046222933038e-07,
+      "loss": 0.4099,
+      "step": 2467
+    },
+    {
+      "epoch": 2.4338180921863444,
+      "grad_norm": 0.1652612101877314,
+      "learning_rate": 8.55808621853299e-07,
+      "loss": 0.4121,
+      "step": 2468
+    },
+    {
+      "epoch": 2.434804042395859,
+      "grad_norm": 0.1558208278498863,
+      "learning_rate": 8.529170559670674e-07,
+      "loss": 0.4186,
+      "step": 2469
+    },
+    {
+      "epoch": 2.4357899926053737,
+      "grad_norm": 0.16267826283367576,
+      "learning_rate": 8.500299277287744e-07,
+      "loss": 0.3965,
+      "step": 2470
+    },
+    {
+      "epoch": 2.436775942814888,
+      "grad_norm": 0.16653164908480583,
+      "learning_rate": 8.47147240227833e-07,
+      "loss": 0.422,
+      "step": 2471
+    },
+    {
+      "epoch": 2.437761893024402,
+      "grad_norm": 0.15965025920038586,
+      "learning_rate": 8.442689965489087e-07,
+      "loss": 0.4152,
+      "step": 2472
+    },
+    {
+      "epoch": 2.4387478432339167,
+      "grad_norm": 0.1634090846806474,
+      "learning_rate": 8.413951997719083e-07,
+      "loss": 0.3924,
+      "step": 2473
+    },
+    {
+      "epoch": 2.4397337934434313,
+      "grad_norm": 0.15968750027676784,
+      "learning_rate": 8.385258529719781e-07,
+      "loss": 0.4201,
+      "step": 2474
+    },
+    {
+      "epoch": 2.4407197436529455,
+      "grad_norm": 0.16285082108863647,
+      "learning_rate": 8.356609592195081e-07,
+      "loss": 0.4077,
+      "step": 2475
+    },
+    {
+      "epoch": 2.4417056938624597,
+      "grad_norm": 0.16079529099764747,
+      "learning_rate": 8.328005215801205e-07,
+      "loss": 0.4261,
+      "step": 2476
+    },
+    {
+      "epoch": 2.4426916440719744,
+      "grad_norm": 0.17260608545522937,
+      "learning_rate": 8.299445431146686e-07,
+      "loss": 0.4062,
+      "step": 2477
+    },
+    {
+      "epoch": 2.443677594281489,
+      "grad_norm": 0.16877987809000522,
+      "learning_rate": 8.270930268792343e-07,
+      "loss": 0.4231,
+      "step": 2478
+    },
+    {
+      "epoch": 2.444663544491003,
+      "grad_norm": 0.1603976054916282,
+      "learning_rate": 8.242459759251259e-07,
+      "loss": 0.405,
+      "step": 2479
+    },
+    {
+      "epoch": 2.4456494947005174,
+      "grad_norm": 0.16335409022031283,
+      "learning_rate": 8.214033932988724e-07,
+      "loss": 0.4152,
+      "step": 2480
+    },
+    {
+      "epoch": 2.446635444910032,
+      "grad_norm": 0.16189146972601887,
+      "learning_rate": 8.185652820422219e-07,
+      "loss": 0.4255,
+      "step": 2481
+    },
+    {
+      "epoch": 2.4476213951195467,
+      "grad_norm": 0.15963137364834074,
+      "learning_rate": 8.15731645192136e-07,
+      "loss": 0.4021,
+      "step": 2482
+    },
+    {
+      "epoch": 2.448607345329061,
+      "grad_norm": 0.16265749135551338,
+      "learning_rate": 8.129024857807943e-07,
+      "loss": 0.437,
+      "step": 2483
+    },
+    {
+      "epoch": 2.449593295538575,
+      "grad_norm": 0.16056692700621705,
+      "learning_rate": 8.100778068355769e-07,
+      "loss": 0.4249,
+      "step": 2484
+    },
+    {
+      "epoch": 2.4505792457480897,
+      "grad_norm": 0.16598876062970291,
+      "learning_rate": 8.072576113790754e-07,
+      "loss": 0.4051,
+      "step": 2485
+    },
+    {
+      "epoch": 2.4515651959576044,
+      "grad_norm": 0.1608088923146115,
+      "learning_rate": 8.0444190242908e-07,
+      "loss": 0.4102,
+      "step": 2486
+    },
+    {
+      "epoch": 2.4525511461671186,
+      "grad_norm": 0.16286213988022255,
+      "learning_rate": 8.016306829985848e-07,
+      "loss": 0.4214,
+      "step": 2487
+    },
+    {
+      "epoch": 2.4535370963766328,
+      "grad_norm": 0.16028060307774927,
+      "learning_rate": 7.988239560957773e-07,
+      "loss": 0.4178,
+      "step": 2488
+    },
+    {
+      "epoch": 2.4545230465861474,
+      "grad_norm": 0.16714415782305406,
+      "learning_rate": 7.960217247240342e-07,
+      "loss": 0.4093,
+      "step": 2489
+    },
+    {
+      "epoch": 2.455508996795662,
+      "grad_norm": 0.16666325491450767,
+      "learning_rate": 7.932239918819262e-07,
+      "loss": 0.4175,
+      "step": 2490
+    },
+    {
+      "epoch": 2.4564949470051762,
+      "grad_norm": 0.17489421957601758,
+      "learning_rate": 7.904307605632111e-07,
+      "loss": 0.437,
+      "step": 2491
+    },
+    {
+      "epoch": 2.4574808972146904,
+      "grad_norm": 0.17206936312624158,
+      "learning_rate": 7.876420337568264e-07,
+      "loss": 0.4237,
+      "step": 2492
+    },
+    {
+      "epoch": 2.458466847424205,
+      "grad_norm": 0.1947580103630649,
+      "learning_rate": 7.848578144468899e-07,
+      "loss": 0.425,
+      "step": 2493
+    },
+    {
+      "epoch": 2.4594527976337197,
+      "grad_norm": 0.15966031523643928,
+      "learning_rate": 7.820781056126986e-07,
+      "loss": 0.4273,
+      "step": 2494
+    },
+    {
+      "epoch": 2.460438747843234,
+      "grad_norm": 0.15855053189533086,
+      "learning_rate": 7.793029102287202e-07,
+      "loss": 0.3993,
+      "step": 2495
+    },
+    {
+      "epoch": 2.461424698052748,
+      "grad_norm": 0.16258951492067503,
+      "learning_rate": 7.76532231264594e-07,
+      "loss": 0.4226,
+      "step": 2496
+    },
+    {
+      "epoch": 2.4624106482622627,
+      "grad_norm": 0.1565035184019167,
+      "learning_rate": 7.73766071685127e-07,
+      "loss": 0.3969,
+      "step": 2497
+    },
+    {
+      "epoch": 2.4633965984717774,
+      "grad_norm": 0.1670879812733264,
+      "learning_rate": 7.710044344502893e-07,
+      "loss": 0.4049,
+      "step": 2498
+    },
+    {
+      "epoch": 2.4643825486812916,
+      "grad_norm": 0.1611325585036167,
+      "learning_rate": 7.682473225152115e-07,
+      "loss": 0.41,
+      "step": 2499
+    },
+    {
+      "epoch": 2.465368498890806,
+      "grad_norm": 0.16312168729718626,
+      "learning_rate": 7.654947388301826e-07,
+      "loss": 0.4094,
+      "step": 2500
+    },
+    {
+      "epoch": 2.4663544491003204,
+      "grad_norm": 0.17710984544089278,
+      "learning_rate": 7.627466863406446e-07,
+      "loss": 0.3929,
+      "step": 2501
+    },
+    {
+      "epoch": 2.467340399309835,
+      "grad_norm": 0.16386566596517432,
+      "learning_rate": 7.600031679871944e-07,
+      "loss": 0.4092,
+      "step": 2502
+    },
+    {
+      "epoch": 2.4683263495193493,
+      "grad_norm": 0.17407210750463106,
+      "learning_rate": 7.572641867055752e-07,
+      "loss": 0.4061,
+      "step": 2503
+    },
+    {
+      "epoch": 2.4693122997288635,
+      "grad_norm": 0.1586376619062199,
+      "learning_rate": 7.54529745426672e-07,
+      "loss": 0.4091,
+      "step": 2504
+    },
+    {
+      "epoch": 2.470298249938378,
+      "grad_norm": 0.20293350660629642,
+      "learning_rate": 7.517998470765142e-07,
+      "loss": 0.4152,
+      "step": 2505
+    },
+    {
+      "epoch": 2.4712842001478927,
+      "grad_norm": 0.15537060243547707,
+      "learning_rate": 7.490744945762729e-07,
+      "loss": 0.414,
+      "step": 2506
+    },
+    {
+      "epoch": 2.472270150357407,
+      "grad_norm": 0.16887866356080697,
+      "learning_rate": 7.463536908422508e-07,
+      "loss": 0.3985,
+      "step": 2507
+    },
+    {
+      "epoch": 2.473256100566921,
+      "grad_norm": 0.16196577013182265,
+      "learning_rate": 7.436374387858863e-07,
+      "loss": 0.4086,
+      "step": 2508
+    },
+    {
+      "epoch": 2.4742420507764358,
+      "grad_norm": 0.1583935835828545,
+      "learning_rate": 7.409257413137411e-07,
+      "loss": 0.3896,
+      "step": 2509
+    },
+    {
+      "epoch": 2.4752280009859504,
+      "grad_norm": 0.16004952218937224,
+      "learning_rate": 7.382186013275117e-07,
+      "loss": 0.4009,
+      "step": 2510
+    },
+    {
+      "epoch": 2.4762139511954646,
+      "grad_norm": 0.16154490144280742,
+      "learning_rate": 7.355160217240114e-07,
+      "loss": 0.4127,
+      "step": 2511
+    },
+    {
+      "epoch": 2.477199901404979,
+      "grad_norm": 0.16446533261954524,
+      "learning_rate": 7.328180053951773e-07,
+      "loss": 0.395,
+      "step": 2512
+    },
+    {
+      "epoch": 2.4781858516144935,
+      "grad_norm": 0.15693037541777302,
+      "learning_rate": 7.301245552280594e-07,
+      "loss": 0.3993,
+      "step": 2513
+    },
+    {
+      "epoch": 2.479171801824008,
+      "grad_norm": 0.1626364085065384,
+      "learning_rate": 7.274356741048283e-07,
+      "loss": 0.4094,
+      "step": 2514
+    },
+    {
+      "epoch": 2.4801577520335223,
+      "grad_norm": 0.16596135636991863,
+      "learning_rate": 7.247513649027582e-07,
+      "loss": 0.4203,
+      "step": 2515
+    },
+    {
+      "epoch": 2.4811437022430365,
+      "grad_norm": 0.20416447092391282,
+      "learning_rate": 7.220716304942349e-07,
+      "loss": 0.4174,
+      "step": 2516
+    },
+    {
+      "epoch": 2.482129652452551,
+      "grad_norm": 0.16086469498638334,
+      "learning_rate": 7.193964737467474e-07,
+      "loss": 0.4303,
+      "step": 2517
+    },
+    {
+      "epoch": 2.4831156026620658,
+      "grad_norm": 0.18584061035614108,
+      "learning_rate": 7.167258975228886e-07,
+      "loss": 0.4186,
+      "step": 2518
+    },
+    {
+      "epoch": 2.48410155287158,
+      "grad_norm": 0.16116384490845706,
+      "learning_rate": 7.140599046803492e-07,
+      "loss": 0.4343,
+      "step": 2519
+    },
+    {
+      "epoch": 2.485087503081094,
+      "grad_norm": 0.16595715949503784,
+      "learning_rate": 7.113984980719107e-07,
+      "loss": 0.416,
+      "step": 2520
+    },
+    {
+      "epoch": 2.486073453290609,
+      "grad_norm": 0.16213255355579434,
+      "learning_rate": 7.08741680545455e-07,
+      "loss": 0.411,
+      "step": 2521
+    },
+    {
+      "epoch": 2.4870594035001234,
+      "grad_norm": 0.5202621030835727,
+      "learning_rate": 7.060894549439474e-07,
+      "loss": 0.4251,
+      "step": 2522
+    },
+    {
+      "epoch": 2.4880453537096376,
+      "grad_norm": 0.1626132249365722,
+      "learning_rate": 7.034418241054414e-07,
+      "loss": 0.4244,
+      "step": 2523
+    },
+    {
+      "epoch": 2.489031303919152,
+      "grad_norm": 0.1658844835442372,
+      "learning_rate": 7.007987908630742e-07,
+      "loss": 0.4194,
+      "step": 2524
+    },
+    {
+      "epoch": 2.4900172541286665,
+      "grad_norm": 0.16676631005694778,
+      "learning_rate": 6.98160358045063e-07,
+      "loss": 0.4097,
+      "step": 2525
+    },
+    {
+      "epoch": 2.491003204338181,
+      "grad_norm": 0.16689683870697927,
+      "learning_rate": 6.955265284747026e-07,
+      "loss": 0.4123,
+      "step": 2526
+    },
+    {
+      "epoch": 2.4919891545476953,
+      "grad_norm": 0.16335327159378088,
+      "learning_rate": 6.928973049703608e-07,
+      "loss": 0.4082,
+      "step": 2527
+    },
+    {
+      "epoch": 2.4929751047572095,
+      "grad_norm": 0.16794880378266772,
+      "learning_rate": 6.902726903454765e-07,
+      "loss": 0.4147,
+      "step": 2528
+    },
+    {
+      "epoch": 2.493961054966724,
+      "grad_norm": 0.16809494987324144,
+      "learning_rate": 6.876526874085609e-07,
+      "loss": 0.4094,
+      "step": 2529
+    },
+    {
+      "epoch": 2.494947005176239,
+      "grad_norm": 0.161228286434843,
+      "learning_rate": 6.850372989631842e-07,
+      "loss": 0.4082,
+      "step": 2530
+    },
+    {
+      "epoch": 2.495932955385753,
+      "grad_norm": 0.16522154052752383,
+      "learning_rate": 6.824265278079834e-07,
+      "loss": 0.4213,
+      "step": 2531
+    },
+    {
+      "epoch": 2.496918905595267,
+      "grad_norm": 0.2903809047831653,
+      "learning_rate": 6.798203767366507e-07,
+      "loss": 0.4135,
+      "step": 2532
+    },
+    {
+      "epoch": 2.497904855804782,
+      "grad_norm": 0.1596744108428839,
+      "learning_rate": 6.7721884853794e-07,
+      "loss": 0.4165,
+      "step": 2533
+    },
+    {
+      "epoch": 2.4988908060142965,
+      "grad_norm": 0.15766713592881945,
+      "learning_rate": 6.746219459956554e-07,
+      "loss": 0.4062,
+      "step": 2534
+    },
+    {
+      "epoch": 2.4998767562238107,
+      "grad_norm": 0.16199793614343647,
+      "learning_rate": 6.720296718886488e-07,
+      "loss": 0.4071,
+      "step": 2535
+    },
+    {
+      "epoch": 2.500862706433325,
+      "grad_norm": 0.16602228019139734,
+      "learning_rate": 6.694420289908215e-07,
+      "loss": 0.4209,
+      "step": 2536
+    },
+    {
+      "epoch": 2.5018486566428395,
+      "grad_norm": 0.15969240055857473,
+      "learning_rate": 6.668590200711222e-07,
+      "loss": 0.4037,
+      "step": 2537
+    },
+    {
+      "epoch": 2.502834606852354,
+      "grad_norm": 0.1598104968992545,
+      "learning_rate": 6.642806478935359e-07,
+      "loss": 0.4161,
+      "step": 2538
+    },
+    {
+      "epoch": 2.5038205570618683,
+      "grad_norm": 0.16839173517332648,
+      "learning_rate": 6.617069152170896e-07,
+      "loss": 0.414,
+      "step": 2539
+    },
+    {
+      "epoch": 2.5048065072713825,
+      "grad_norm": 0.163374646034395,
+      "learning_rate": 6.591378247958435e-07,
+      "loss": 0.4239,
+      "step": 2540
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3042,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 254,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.802771677524787e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}